Last active
August 8, 2026 20:46
-
-
Save sshh12/42fc789e03aec6aa6b9ef841f66cabcc to your computer and use it in GitHub Desktop.
Exploring Claude/GPT Knowledge Cutoffs
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| <!doctype html> | |
| <meta charset="utf-8"> | |
| <title>Identity vintage - which era does each model think it is from?</title> | |
| <style> | |
| body { background:#f9f9f7; color:#0b0b0b; margin:0; | |
| font:13px/1.45 system-ui, "Segoe UI", sans-serif; } | |
| .wrap { max-width:1500px; margin:24px auto 60px; padding:0 18px; } | |
| h1 { font-size:20px; margin:0 0 6px; } | |
| .sub { color:#52514e; max-width:960px; margin:0 0 14px; } | |
| .modes { margin:0 0 10px; } | |
| .modes button { font:12px system-ui; padding:4px 10px; margin-right:6px; | |
| border:1px solid #c3c2b7; background:#fcfcfb; color:#52514e; | |
| border-radius:4px; cursor:pointer; } | |
| .modes button.on { background:#0b0b0b; color:#fcfcfb; border-color:#0b0b0b; } | |
| .legend { color:#52514e; font-size:11.5px; margin:0 0 14px; } | |
| .legend span { margin-right:14px; white-space:nowrap; } | |
| .chip { display:inline-block; width:18px; height:9px; border-radius:2px; | |
| vertical-align:-1px; } | |
| #chart { position:relative; background:#fcfcfb; border:1px solid #e1e0d9; | |
| border-radius:6px; overflow-x:auto; } | |
| #cv { position:relative; } | |
| .grid { position:absolute; top:0; bottom:0; width:0; | |
| border-left:1px solid #eeede7; } | |
| .grid.major { border-left-color:#e1e0d9; } | |
| .grid.zero { border-left:1.5px dashed #898781; } | |
| .grid.today { border-left:1.5px dashed #b23b2e; } | |
| .tick { position:absolute; font-size:10px; color:#898781; | |
| transform:translateX(-50%); } | |
| .hdr { position:absolute; font-size:11px; font-weight:600; color:#52514e; | |
| letter-spacing:.06em; } | |
| .hdr b { color:#0b0b0b; } | |
| .rowlab { position:absolute; left:0; width:196px; padding:5px 8px 0 12px; | |
| background:#fcfcfb; z-index:5; } | |
| .rowlab .nm { font-size:11.5px; font-weight:600; white-space:nowrap; } | |
| .rowlab .st { font-size:9.5px; color:#898781; white-space:nowrap; } | |
| .band { position:absolute; height:16px; top:8px; border-radius:3px; } | |
| .band.clipL { border-left:2px dashed rgba(11,11,11,.35); | |
| border-top-left-radius:0; border-bottom-left-radius:0; } | |
| .bandlab { position:absolute; top:9.5px; font-size:9px; color:#fcfcfb; | |
| white-space:nowrap; overflow:hidden; text-shadow:0 0 2px rgba(0,0,0,.25); | |
| pointer-events:none; } | |
| .mid { position:absolute; top:4px; width:2px; height:38px; background:#1baf7a; } | |
| .rel { position:absolute; top:32px; width:0; height:0; | |
| border-left:5px solid transparent; border-right:5px solid transparent; | |
| border-bottom:8px solid; transform:translateX(-5px); } | |
| .vint { position:absolute; top:11px; width:9px; height:9px; | |
| background:#4d3492; border:1.5px solid #fcfcfb; | |
| transform:translateX(-6px) rotate(45deg); z-index:4; | |
| box-shadow:0 0 0 1px rgba(77,52,146,.35); } | |
| .lagline { position:absolute; top:38px; border-top:1.5px dotted #7a5fc7; } | |
| .rline { position:absolute; left:196px; right:0; border-top:1px solid #f0efe9; } | |
| #tip { position:fixed; display:none; background:#0b0b0b; color:#f6f5f1; | |
| font-size:11px; line-height:1.5; padding:7px 10px; border-radius:5px; | |
| max-width:360px; z-index:50; pointer-events:none; } | |
| #tip b { color:#fff; } | |
| details { margin-top:18px; color:#52514e; } | |
| details table { border-collapse:collapse; font-size:11px; margin-top:8px; } | |
| details td { border:1px solid #e1e0d9; padding:2px 8px; } | |
| summary { cursor:pointer; } | |
| h3.sk { font-size:13px; margin:18px 0 6px; color:#52514e; | |
| letter-spacing:.05em; text-transform:uppercase; } | |
| path.lk { opacity:.42; stroke:#fcfcfb; stroke-width:.7; } | |
| path.lk:hover { opacity:.75; } | |
| path.lk.dim { opacity:.05; } | |
| rect.dim { opacity:.22; } | |
| text.dim { opacity:.15; } | |
| .skwrap { position:relative; display:inline-block; } | |
| .callout { position:absolute; font:italic 10.5px/1.35 system-ui; | |
| color:#52514e; background:rgba(252,252,251,.92); | |
| padding:3px 8px; border-left:2px solid #c3c2b7; | |
| border-radius:0 3px 3px 0; max-width:190px; pointer-events:none; } | |
| .callout.r { border-left:none; border-right:2px solid #c3c2b7; | |
| border-radius:3px 0 0 3px; text-align:right; } | |
| </style> | |
| <div class="wrap"> | |
| <h1>Identity vintage: which era does each model think it is from?</h1> | |
| <p class="sub">Join of the identity probe (“what model are you?”, 50 rows/model, | |
| no system prompt) with release-date metadata. A versioned self-claim is a timestamp: | |
| the claimed name was <i>the</i> current model only between its release and the release | |
| of the next better model in that line, so each claim maps to that | |
| <b>currency era</b> (band; opacity = share of the 50 replies). The | |
| <span style="color:#4d3492">◆ identity vintage</span> is the claim-weighted | |
| middle of those eras — an estimate of when the model’s identity-forming | |
| training data was written — shown against the | |
| <span style="color:#1baf7a">| measured knowledge end</span> and the model’s own | |
| <b>▲ release</b>. Unversioned claims (“Claude”, “ChatGPT”) carry no date and are | |
| excluded; eras assume identity text concentrates while a name is current, so | |
| vintages are early-biased where a name lingered (e.g. GPT-4).</p> | |
| <div class="modes"> | |
| <button id="bcal" class="on">Calendar</button> | |
| <button id="brel">Center: release date</button> | |
| </div> | |
| <p class="legend"> | |
| <span><span class="chip" style="background:rgba(42,120,214,.75)"></span> claimed OpenAI-name era</span> | |
| <span><span class="chip" style="background:rgba(235,104,52,.75)"></span> claimed Anthropic-name era</span> | |
| <span><span class="chip" style="background:rgba(107,114,128,.75)"></span> other maker</span> | |
| <span style="color:#4d3492">◆ identity vintage (claim-weighted)</span> | |
| <span style="color:#1baf7a">| measured knowledge end</span> | |
| <span>▲ release date</span> | |
| <span style="color:#b23b2e">┊ probe run (2026-08-07)</span> | |
| </p> | |
| <div id="chart"><div id="cv"></div></div> | |
| <h2 style="font-size:16px;margin:28px 0 4px">Identity vintage vs fact-based knowledge end</h2> | |
| <p class="sub">One point per model, axes matching the implied-today scatter: | |
| <b>x = fact-based knowledge end</b> (measured midpoint), <b>y = identity vintage</b> | |
| (claim-weighted era midpoint from above). Models with no measurable fact-based estimate | |
| (flat accuracy curve in the bench window — the GPT-4/4o/4.1/o-series controls) are | |
| excluded, as are <b>gpt-5 and gpt-5.1</b>: gpt-5.1’s 23-month-wide accuracy | |
| transition yields no usable midpoint, and gpt-5’s identity is set by | |
| post-training-era data (it claims “GPT-4.1”, a name that postdates its corpus) | |
| — the one case where identity vintage sits <i>above</i> factual knowledge. | |
| Points below the dashed y = x line describe themselves as an older model | |
| than their knowledge says they are; a <b>dashed ring</b> marks the remaining models | |
| whose transition is ≥12 months wide (noisy x).</p> | |
| <div id="scatter"></div> | |
| <h2 style="font-size:16px;margin:28px 0 4px">Identity inheritance: which older model each model thinks it is</h2> | |
| <p class="sub">Multilayer Sankey per lab: every model (and every claimed-only legacy name) | |
| is a node in its generation column — oldest generations left, newest right | |
| (GPT-5.6 Sol/Terra/Luna; Fable/Sonnet/Opus 5). A ribbon flows from an older node into | |
| each newer model that claims to <i>be</i> it; width = number of that model’s 50 | |
| identity replies. Mid-generation nodes both receive their own claims (left edge) and | |
| emit the claims newer models make about them (right edge), so identity visibly | |
| propagates — and skips — through generations. <b>Ribbon shade = era of the | |
| claimed identity</b>: pale = oldest generations, dark = newest, so long pale throws are | |
| stale inherited identity and short dark hops are fresh self-knowledge. | |
| <b>Hover any node to isolate its lineage.</b> Unversioned claims | |
| (“Claude”, “ChatGPT”) are excluded here; tooltips carry the counts.</p> | |
| <h3 class="sk">OpenAI</h3><div id="sk-openai"></div> | |
| <h3 class="sk">Anthropic</h3><div id="sk-anthropic"></div> | |
| <details> | |
| <summary>Method: claim → era mapping table and assumptions</summary> | |
| <p>Era start = first public availability of the claimed name (announcement/GA); | |
| era end = release of the next better model in the same product line (bench | |
| registry where available, public dates otherwise). <b>Identity vintage</b> = | |
| claim-count-weighted mean of per-claim date estimates under a calibrated era | |
| model: 70% of a claim’s mention mass is uniform within the era, 30% sits in a | |
| post-supersession tail centered +15 months after the successor ships (names keep | |
| being written about after they are superseded — the “GPT-4 effect”). | |
| Cross-provider claims (a Claude answering “GPT-4”) are drawn everywhere but | |
| excluded from the vintage point estimate: as generic fame, they carry almost no | |
| date information. Parameters were fit by <code>fit_vintage.py</code> to minimize the | |
| median |vintage − measured knowledge end| over the 21 measured models | |
| (median residual −6.1 → −0.2 mo, median abs 6.4 → 2.2 mo; the | |
| centering is calibrated, so read the scatter for its residual spread, not its | |
| mean). Models whose replies are all unversioned get no vintage.</p> | |
| <table><tr><td><b>claimed name</b></td><td><b>era start</b></td><td><b>era end</b></td><td><b>why it ends</b></td></tr> | |
| <tr><td>Claude 2</td><td>2023-07-11</td><td>2024-03-04</td><td>superseded by Claude 3</td></tr><tr><td>Claude 3</td><td>2024-03-04</td><td>2024-06-20</td><td>superseded by Claude 3.5 Sonnet</td></tr><tr><td>Claude 3 Haiku</td><td>2024-03-13</td><td>2024-11-04</td><td>superseded by Claude 3.5 Haiku</td></tr><tr><td>Claude 3 Opus</td><td>2024-03-04</td><td>2024-06-20</td><td>superseded by Claude 3.5 Sonnet</td></tr><tr><td>Claude 3 Sonnet</td><td>2024-03-04</td><td>2024-06-20</td><td>superseded by Claude 3.5 Sonnet</td></tr><tr><td>Claude 3.5</td><td>2024-06-20</td><td>2025-02-24</td><td>superseded by Claude 3.7 Sonnet</td></tr><tr><td>Claude 3.5 Haiku</td><td>2024-11-04</td><td>2025-10-15</td><td>superseded by Claude Haiku 4.5</td></tr><tr><td>Claude 3.5 Sonnet</td><td>2024-06-20</td><td>2025-02-24</td><td>superseded by Claude 3.7 Sonnet</td></tr><tr><td>Claude 3.7 Sonnet</td><td>2025-02-24</td><td>2025-05-22</td><td>superseded by Claude Sonnet 4</td></tr><tr><td>Claude 4</td><td>2025-05-22</td><td>2025-09-29</td><td>superseded by Claude Sonnet 4.5</td></tr><tr><td>Claude Opus 4.1</td><td>2025-08-05</td><td>2025-11-24</td><td>superseded by Claude Opus 4.5</td></tr><tr><td>Claude Opus 4.5</td><td>2025-11-24</td><td>2026-02-05</td><td>superseded by Claude Opus 4.6</td></tr><tr><td>Claude Sonnet 3.5</td><td>2024-06-20</td><td>2025-02-24</td><td>superseded by Claude 3.7 Sonnet</td></tr><tr><td>Claude Sonnet 3.7</td><td>2025-02-24</td><td>2025-05-22</td><td>superseded by Claude Sonnet 4</td></tr><tr><td>Claude Sonnet 4</td><td>2025-05-22</td><td>2025-09-29</td><td>superseded by Claude Sonnet 4.5</td></tr><tr><td>Claude Sonnet 4.5</td><td>2025-09-29</td><td>2026-02-17</td><td>superseded by Claude Sonnet 4.6</td></tr><tr><td>GPT-3</td><td>2020-06-11</td><td>2022-11-30</td><td>superseded by ChatGPT (GPT-3.5)</td></tr><tr><td>GPT-3.5</td><td>2022-11-30</td><td>2023-03-14</td><td>superseded by GPT-4</td></tr><tr><td>GPT-3.5 Turbo</td><td>2022-11-30</td><td>2023-03-14</td><td>superseded by GPT-4</td></tr><tr><td>GPT-4</td><td>2023-03-14</td><td>2023-11-06</td><td>superseded by GPT-4 Turbo</td></tr><tr><td>GPT-4 Turbo</td><td>2023-11-06</td><td>2024-05-13</td><td>superseded by GPT-4o</td></tr><tr><td>GPT-4.1</td><td>2025-04-14</td><td>2025-08-07</td><td>superseded by GPT-5</td></tr><tr><td>GPT-4.1 mini</td><td>2025-04-14</td><td>2025-08-07</td><td>superseded by GPT-5 mini</td></tr><tr><td>GPT-4.5</td><td>2025-02-27</td><td>2025-08-07</td><td>superseded by GPT-5</td></tr><tr><td>GPT-4o</td><td>2024-05-13</td><td>2025-04-14</td><td>superseded by GPT-4.1</td></tr><tr><td>GPT-4o mini</td><td>2024-07-18</td><td>2025-04-14</td><td>superseded by GPT-4.1 mini</td></tr><tr><td>GPT-5</td><td>2025-08-07</td><td>2025-11-13</td><td>superseded by GPT-5.1</td></tr><tr><td>GPT-5 mini</td><td>2025-08-07</td><td>2026-03-17</td><td>superseded by GPT-5.4 mini</td></tr><tr><td>GPT-5.1</td><td>2025-11-13</td><td>2025-12-11</td><td>superseded by GPT-5.2</td></tr><tr><td>GPT-5.2</td><td>2025-12-11</td><td>2026-03-05</td><td>superseded by GPT-5.4</td></tr><tr><td>Llama 3</td><td>2024-04-18</td><td>2024-07-23</td><td>superseded by Llama 3.1</td></tr><tr><td>o1</td><td>2024-09-12</td><td>2025-04-16</td><td>superseded by o3</td></tr><tr><td>o3</td><td>2025-04-16</td><td>2025-08-07</td><td>superseded by GPT-5</td></tr><tr><td>o3-mini</td><td>2025-01-31</td><td>2025-04-16</td><td>superseded by o4-mini</td></tr><tr><td>o3-pro</td><td>2025-06-10</td><td>2025-08-07</td><td>superseded by GPT-5 pro</td></tr><tr><td>o4-mini</td><td>2025-04-16</td><td>2025-08-07</td><td>superseded by GPT-5 mini</td></tr> | |
| </table> | |
| </details> | |
| </div> | |
| <div id="tip"></div> | |
| <script> | |
| var B = {"models":[{"id":"gpt-4","name":"GPT-4","provider":"openai","release":"2023-03-14","published":"2021-09","midpoint":null,"n":50,"claims":[{"claim":"GPT-3","count":40,"share":0.8,"start":"2020-06-11","end":"2022-11-30","succ":"ChatGPT (GPT-3.5)","maker":"openai"}],"unmapped":[{"claim":"(none)","count":10}],"mapped_n":40,"wide":false,"xbroken":false,"vintage":"2022-06","lag_rel":-9.3,"lag_mid":null},{"id":"gpt-4-turbo","name":"GPT-4 Turbo","provider":"openai","release":"2024-04-09","published":"2023-12","midpoint":null,"n":50,"claims":[{"claim":"GPT-4","count":30,"share":0.6,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"GPT-3","count":13,"share":0.26,"start":"2020-06-11","end":"2022-11-30","succ":"ChatGPT (GPT-3.5)","maker":"openai"},{"claim":"GPT-3.5","count":3,"share":0.06,"start":"2022-11-30","end":"2023-03-14","succ":"GPT-4","maker":"openai"}],"unmapped":[{"claim":"ChatGPT","count":2},{"claim":"GPT (unversioned)","count":2}],"mapped_n":46,"wide":false,"xbroken":false,"vintage":"2023-07","lag_rel":-9.0,"lag_mid":null},{"id":"gpt-4o","name":"GPT-4o","provider":"openai","release":"2024-05-13","published":"2023-10","midpoint":null,"n":50,"claims":[{"claim":"GPT-4","count":34,"share":0.68,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"GPT-3.5","count":7,"share":0.14,"start":"2022-11-30","end":"2023-03-14","succ":"GPT-4","maker":"openai"},{"claim":"GPT-3.5 Turbo","count":5,"share":0.1,"start":"2022-11-30","end":"2023-03-14","succ":"GPT-4","maker":"openai"},{"claim":"GPT-4 Turbo","count":3,"share":0.06,"start":"2023-11-06","end":"2024-05-13","succ":"GPT-4o","maker":"openai"}],"unmapped":[{"claim":"ChatGPT","count":1}],"mapped_n":49,"wide":false,"xbroken":false,"vintage":"2023-12","lag_rel":-5.6,"lag_mid":null},{"id":"gpt-4o-mini","name":"GPT-4o mini","provider":"openai","release":"2024-07-18","published":"2023-10","midpoint":null,"n":50,"claims":[{"claim":"GPT-3.5","count":22,"share":0.44,"start":"2022-11-30","end":"2023-03-14","succ":"GPT-4","maker":"openai"},{"claim":"GPT-4","count":10,"share":0.2,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"GPT-3","count":8,"share":0.16,"start":"2020-06-11","end":"2022-11-30","succ":"ChatGPT (GPT-3.5)","maker":"openai"},{"claim":"GPT-3.5 Turbo","count":3,"share":0.06,"start":"2022-11-30","end":"2023-03-14","succ":"GPT-4","maker":"openai"},{"claim":"GPT-4 Turbo","count":1,"share":0.02,"start":"2023-11-06","end":"2024-05-13","succ":"GPT-4o","maker":"openai"}],"unmapped":[{"claim":"ChatGPT","count":6}],"mapped_n":44,"wide":false,"xbroken":false,"vintage":"2023-06","lag_rel":-13.4,"lag_mid":null},{"id":"o3-mini","name":"o3-mini","provider":"openai","release":"2025-01-31","published":"2023-10","midpoint":null,"n":50,"claims":[{"claim":"GPT-4","count":46,"share":0.92,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"GPT-4 Turbo","count":1,"share":0.02,"start":"2023-11-06","end":"2024-05-13","succ":"GPT-4o","maker":"openai"}],"unmapped":[{"claim":"ChatGPT","count":3}],"mapped_n":47,"wide":false,"xbroken":false,"vintage":"2024-01","lag_rel":-12.9,"lag_mid":null},{"id":"gpt-4.1","name":"GPT-4.1","provider":"openai","release":"2025-04-14","published":"2024-06","midpoint":null,"n":50,"claims":[{"claim":"GPT-4","count":43,"share":0.86,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"GPT-4o","count":6,"share":0.12,"start":"2024-05-13","end":"2025-04-14","succ":"GPT-4.1","maker":"openai"},{"claim":"GPT-4 Turbo","count":1,"share":0.02,"start":"2023-11-06","end":"2024-05-13","succ":"GPT-4o","maker":"openai"}],"unmapped":[],"mapped_n":50,"wide":false,"xbroken":false,"vintage":"2024-03","lag_rel":-13.4,"lag_mid":null},{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","provider":"openai","release":"2025-04-14","published":"2024-06","midpoint":null,"n":50,"claims":[{"claim":"GPT-4","count":40,"share":0.8,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"GPT-4o mini","count":5,"share":0.1,"start":"2024-07-18","end":"2025-04-14","succ":"GPT-4.1 mini","maker":"openai"},{"claim":"GPT-4o","count":3,"share":0.06,"start":"2024-05-13","end":"2025-04-14","succ":"GPT-4.1","maker":"openai"},{"claim":"GPT-4 Turbo","count":1,"share":0.02,"start":"2023-11-06","end":"2024-05-13","succ":"GPT-4o","maker":"openai"}],"unmapped":[{"claim":"GPT-4 Chat preview","count":1}],"mapped_n":49,"wide":false,"xbroken":false,"vintage":"2024-04","lag_rel":-12.6,"lag_mid":null},{"id":"gpt-4.1-nano","name":"GPT-4.1 nano","provider":"openai","release":"2025-04-14","published":"2024-06","midpoint":null,"n":50,"claims":[{"claim":"GPT-4","count":50,"share":1.0,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"}],"unmapped":[],"mapped_n":50,"wide":false,"xbroken":false,"vintage":"2024-01","lag_rel":-15.5,"lag_mid":null},{"id":"o3","name":"o3","provider":"openai","release":"2025-04-16","published":"2024-06","midpoint":null,"n":50,"claims":[{"claim":"GPT-4","count":32,"share":0.64,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"GPT-4o","count":10,"share":0.2,"start":"2024-05-13","end":"2025-04-14","succ":"GPT-4.1","maker":"openai"},{"claim":"GPT-4o mini","count":3,"share":0.06,"start":"2024-07-18","end":"2025-04-14","succ":"GPT-4.1 mini","maker":"openai"}],"unmapped":[{"claim":"(none)","count":4},{"claim":"ChatGPT","count":1}],"mapped_n":45,"wide":false,"xbroken":false,"vintage":"2024-06","lag_rel":-10.8,"lag_mid":null},{"id":"o4-mini","name":"o4-mini","provider":"openai","release":"2025-04-16","published":"2024-06","midpoint":null,"n":50,"claims":[{"claim":"GPT-4","count":33,"share":0.66,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"GPT-4 Turbo","count":8,"share":0.16,"start":"2023-11-06","end":"2024-05-13","succ":"GPT-4o","maker":"openai"},{"claim":"GPT-4o mini","count":1,"share":0.02,"start":"2024-07-18","end":"2025-04-14","succ":"GPT-4.1 mini","maker":"openai"},{"claim":"GPT-3.5 Turbo","count":1,"share":0.02,"start":"2022-11-30","end":"2023-03-14","succ":"GPT-4","maker":"openai"}],"unmapped":[{"claim":"(none)","count":4},{"claim":"ChatGPT","count":3}],"mapped_n":43,"wide":false,"xbroken":false,"vintage":"2024-02","lag_rel":-14.0,"lag_mid":null},{"id":"gpt-5","name":"GPT-5","provider":"openai","release":"2025-08-07","published":"2024-09","midpoint":"2024-06-28","n":50,"claims":[{"claim":"GPT-4o mini","count":17,"share":0.34,"start":"2024-07-18","end":"2025-04-14","succ":"GPT-4.1 mini","maker":"openai"},{"claim":"GPT-4o","count":14,"share":0.28,"start":"2024-05-13","end":"2025-04-14","succ":"GPT-4.1","maker":"openai"},{"claim":"GPT-4","count":9,"share":0.18,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"GPT-4.1","count":5,"share":0.1,"start":"2025-04-14","end":"2025-08-07","succ":"GPT-5","maker":"openai"},{"claim":"GPT-4.1 mini","count":4,"share":0.08,"start":"2025-04-14","end":"2025-08-07","succ":"GPT-5 mini","maker":"openai"},{"claim":"o4-mini","count":1,"share":0.02,"start":"2025-04-16","end":"2025-08-07","succ":"GPT-5 mini","maker":"openai"}],"unmapped":[],"mapped_n":50,"wide":false,"xbroken":true,"vintage":"2025-04","lag_rel":-4.5,"lag_mid":8.8},{"id":"gpt-5-mini","name":"GPT-5 mini","provider":"openai","release":"2025-08-07","published":"2024-05","midpoint":null,"n":50,"claims":[{"claim":"GPT-4o mini","count":22,"share":0.44,"start":"2024-07-18","end":"2025-04-14","succ":"GPT-4.1 mini","maker":"openai"},{"claim":"GPT-4o","count":20,"share":0.4,"start":"2024-05-13","end":"2025-04-14","succ":"GPT-4.1","maker":"openai"},{"claim":"GPT-4","count":6,"share":0.12,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"GPT-4.1","count":1,"share":0.02,"start":"2025-04-14","end":"2025-08-07","succ":"GPT-5","maker":"openai"}],"unmapped":[{"claim":"GPT-4o preview","count":1}],"mapped_n":49,"wide":false,"xbroken":false,"vintage":"2025-04","lag_rel":-4.6,"lag_mid":null},{"id":"gpt-5-nano","name":"GPT-5 nano","provider":"openai","release":"2025-08-07","published":"2024-05","midpoint":null,"n":50,"claims":[{"claim":"GPT-4","count":31,"share":0.62,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"GPT-4 Turbo","count":11,"share":0.22,"start":"2023-11-06","end":"2024-05-13","succ":"GPT-4o","maker":"openai"},{"claim":"GPT-4o","count":7,"share":0.14,"start":"2024-05-13","end":"2025-04-14","succ":"GPT-4.1","maker":"openai"}],"unmapped":[{"claim":"ChatGPT","count":1}],"mapped_n":49,"wide":false,"xbroken":false,"vintage":"2024-05","lag_rel":-15.4,"lag_mid":null},{"id":"gpt-5.1","name":"GPT-5.1","provider":"openai","release":"2025-11-13","published":"2024-09","midpoint":"2024-09-05","n":50,"claims":[{"claim":"GPT-4.1","count":34,"share":0.68,"start":"2025-04-14","end":"2025-08-07","succ":"GPT-5","maker":"openai"},{"claim":"GPT-4.1 mini","count":8,"share":0.16,"start":"2025-04-14","end":"2025-08-07","succ":"GPT-5 mini","maker":"openai"},{"claim":"GPT-4o","count":4,"share":0.08,"start":"2024-05-13","end":"2025-04-14","succ":"GPT-4.1","maker":"openai"},{"claim":"GPT-4","count":2,"share":0.04,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"o3-mini","count":1,"share":0.02,"start":"2025-01-31","end":"2025-04-16","succ":"o4-mini","maker":"openai"}],"unmapped":[{"claim":"o4","count":1}],"mapped_n":49,"wide":true,"xbroken":true,"vintage":"2025-10","lag_rel":-1.5,"lag_mid":12.8},{"id":"gpt-5.2","name":"GPT-5.2","provider":"openai","release":"2025-12-11","published":"2025-08","midpoint":"2025-08-12","n":50,"claims":[{"claim":"GPT-4","count":17,"share":0.34,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"GPT-4.1","count":13,"share":0.26,"start":"2025-04-14","end":"2025-08-07","succ":"GPT-5","maker":"openai"},{"claim":"GPT-4o","count":6,"share":0.12,"start":"2024-05-13","end":"2025-04-14","succ":"GPT-4.1","maker":"openai"},{"claim":"GPT-5","count":4,"share":0.08,"start":"2025-08-07","end":"2025-11-13","succ":"GPT-5.1","maker":"openai"},{"claim":"o4-mini","count":2,"share":0.04,"start":"2025-04-16","end":"2025-08-07","succ":"GPT-5 mini","maker":"openai"}],"unmapped":[{"claim":"(none)","count":7},{"claim":"ChatGPT","count":1}],"mapped_n":42,"wide":false,"xbroken":false,"vintage":"2025-02","lag_rel":-10.6,"lag_mid":-6.6},{"id":"gpt-5.3-codex","name":"GPT-5.3-Codex","provider":"openai","release":"2026-02-05","published":"2025-08","midpoint":"2025-08-13","n":50,"claims":[{"claim":"o4-mini","count":22,"share":0.44,"start":"2025-04-16","end":"2025-08-07","succ":"GPT-5 mini","maker":"openai"},{"claim":"GPT-5","count":8,"share":0.16,"start":"2025-08-07","end":"2025-11-13","succ":"GPT-5.1","maker":"openai"},{"claim":"GPT-4o","count":5,"share":0.1,"start":"2024-05-13","end":"2025-04-14","succ":"GPT-4.1","maker":"openai"},{"claim":"GPT-4.1","count":4,"share":0.08,"start":"2025-04-14","end":"2025-08-07","succ":"GPT-5","maker":"openai"},{"claim":"o3","count":4,"share":0.08,"start":"2025-04-16","end":"2025-08-07","succ":"GPT-5","maker":"openai"},{"claim":"GPT-4","count":4,"share":0.08,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"}],"unmapped":[{"claim":"ChatGPT","count":2},{"claim":"GPT (unversioned)","count":1}],"mapped_n":47,"wide":false,"xbroken":false,"vintage":"2025-09","lag_rel":-4.7,"lag_mid":1.0},{"id":"gpt-5.4","name":"GPT-5.4","provider":"openai","release":"2026-03-05","published":"2025-08","midpoint":"2025-08-21","n":50,"claims":[{"claim":"GPT-5","count":24,"share":0.48,"start":"2025-08-07","end":"2025-11-13","succ":"GPT-5.1","maker":"openai"},{"claim":"o3","count":13,"share":0.26,"start":"2025-04-16","end":"2025-08-07","succ":"GPT-5","maker":"openai"},{"claim":"o4-mini","count":5,"share":0.1,"start":"2025-04-16","end":"2025-08-07","succ":"GPT-5 mini","maker":"openai"},{"claim":"GPT-4.1","count":2,"share":0.04,"start":"2025-04-14","end":"2025-08-07","succ":"GPT-5","maker":"openai"},{"claim":"GPT-5.1","count":1,"share":0.02,"start":"2025-11-13","end":"2025-12-11","succ":"GPT-5.2","maker":"openai"},{"claim":"GPT-4o","count":1,"share":0.02,"start":"2024-05-13","end":"2025-04-14","succ":"GPT-4.1","maker":"openai"}],"unmapped":[{"claim":"ChatGPT","count":4}],"mapped_n":46,"wide":false,"xbroken":false,"vintage":"2026-01","lag_rel":-2.0,"lag_mid":4.5},{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","provider":"openai","release":"2026-03-17","published":"2025-08","midpoint":"2025-12-10","n":50,"claims":[{"claim":"o4-mini","count":11,"share":0.22,"start":"2025-04-16","end":"2025-08-07","succ":"GPT-5 mini","maker":"openai"},{"claim":"GPT-4","count":9,"share":0.18,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"GPT-4o","count":9,"share":0.18,"start":"2024-05-13","end":"2025-04-14","succ":"GPT-4.1","maker":"openai"},{"claim":"GPT-4.1","count":8,"share":0.16,"start":"2025-04-14","end":"2025-08-07","succ":"GPT-5","maker":"openai"},{"claim":"GPT-4o mini","count":4,"share":0.08,"start":"2024-07-18","end":"2025-04-14","succ":"GPT-4.1 mini","maker":"openai"}],"unmapped":[{"claim":"ChatGPT","count":9}],"mapped_n":41,"wide":false,"xbroken":false,"vintage":"2025-05","lag_rel":-11.0,"lag_mid":-7.8},{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","provider":"openai","release":"2026-03-17","published":"2025-08","midpoint":"2025-02-23","n":50,"claims":[{"claim":"GPT-4","count":12,"share":0.24,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"GPT-4o","count":12,"share":0.24,"start":"2024-05-13","end":"2025-04-14","succ":"GPT-4.1","maker":"openai"},{"claim":"GPT-4o mini","count":10,"share":0.2,"start":"2024-07-18","end":"2025-04-14","succ":"GPT-4.1 mini","maker":"openai"},{"claim":"GPT-4.1","count":6,"share":0.12,"start":"2025-04-14","end":"2025-08-07","succ":"GPT-5","maker":"openai"},{"claim":"GPT-4.1 mini","count":3,"share":0.06,"start":"2025-04-14","end":"2025-08-07","succ":"GPT-5 mini","maker":"openai"}],"unmapped":[{"claim":"ChatGPT","count":6},{"claim":"GPT (unversioned)","count":1}],"mapped_n":43,"wide":false,"xbroken":false,"vintage":"2025-02","lag_rel":-13.5,"lag_mid":-0.7},{"id":"gpt-5.5","name":"GPT-5.5","provider":"openai","release":"2026-04-23","published":"2025-12","midpoint":"2025-12-14","n":50,"claims":[{"claim":"GPT-4o","count":24,"share":0.48,"start":"2024-05-13","end":"2025-04-14","succ":"GPT-4.1","maker":"openai"},{"claim":"GPT-5","count":9,"share":0.18,"start":"2025-08-07","end":"2025-11-13","succ":"GPT-5.1","maker":"openai"}],"unmapped":[{"claim":"ChatGPT","count":15},{"claim":"GPT (unversioned)","count":2}],"mapped_n":33,"wide":false,"xbroken":false,"vintage":"2025-08","lag_rel":-9.0,"lag_mid":-4.7},{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","provider":"openai","release":"2026-07-09","published":"2026-02","midpoint":"2026-04-17","n":50,"claims":[{"claim":"GPT-5","count":22,"share":0.44,"start":"2025-08-07","end":"2025-11-13","succ":"GPT-5.1","maker":"openai"},{"claim":"o3-mini","count":7,"share":0.14,"start":"2025-01-31","end":"2025-04-16","succ":"o4-mini","maker":"openai"},{"claim":"GPT-5.2","count":4,"share":0.08,"start":"2025-12-11","end":"2026-03-05","succ":"GPT-5.4","maker":"openai"},{"claim":"o3","count":2,"share":0.04,"start":"2025-04-16","end":"2025-08-07","succ":"GPT-5","maker":"openai"},{"claim":"GPT-5 mini","count":2,"share":0.04,"start":"2025-08-07","end":"2026-03-17","succ":"GPT-5.4 mini","maker":"openai"},{"claim":"o3-pro","count":1,"share":0.02,"start":"2025-06-10","end":"2025-08-07","succ":"GPT-5 pro","maker":"openai"}],"unmapped":[{"claim":"ChatGPT","count":12}],"mapped_n":38,"wide":true,"xbroken":false,"vintage":"2026-02","lag_rel":-5.4,"lag_mid":-2.7},{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","provider":"openai","release":"2026-07-09","published":"2026-02","midpoint":"2026-03-05","n":50,"claims":[{"claim":"GPT-5","count":13,"share":0.26,"start":"2025-08-07","end":"2025-11-13","succ":"GPT-5.1","maker":"openai"},{"claim":"o3","count":7,"share":0.14,"start":"2025-04-16","end":"2025-08-07","succ":"GPT-5","maker":"openai"},{"claim":"GPT-5.2","count":2,"share":0.04,"start":"2025-12-11","end":"2026-03-05","succ":"GPT-5.4","maker":"openai"}],"unmapped":[{"claim":"ChatGPT","count":19},{"claim":"GPT (unversioned)","count":9}],"mapped_n":22,"wide":false,"xbroken":false,"vintage":"2026-02","lag_rel":-5.2,"lag_mid":-1.1},{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","provider":"openai","release":"2026-07-09","published":"2026-02","midpoint":"2026-02-12","n":50,"claims":[{"claim":"GPT-5","count":38,"share":0.76,"start":"2025-08-07","end":"2025-11-13","succ":"GPT-5.1","maker":"openai"},{"claim":"GPT-4","count":1,"share":0.02,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"}],"unmapped":[{"claim":"ChatGPT","count":11}],"mapped_n":39,"wide":false,"xbroken":false,"vintage":"2026-02","lag_rel":-5.1,"lag_mid":-0.2},{"id":"claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5","provider":"anthropic","release":"2025-09-29","published":"2025-01","midpoint":"2025-02-20","n":50,"claims":[{"claim":"Claude 3.5 Sonnet","count":31,"share":0.62,"start":"2024-06-20","end":"2025-02-24","succ":"Claude 3.7 Sonnet","maker":"anthropic"},{"claim":"GPT-4","count":10,"share":0.2,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"Claude 3.5","count":4,"share":0.08,"start":"2024-06-20","end":"2025-02-24","succ":"Claude 3.7 Sonnet","maker":"anthropic"},{"claim":"Claude 3.7 Sonnet","count":2,"share":0.04,"start":"2025-02-24","end":"2025-05-22","succ":"Claude Sonnet 4","maker":"anthropic"}],"unmapped":[{"claim":"Claude","count":3}],"mapped_n":47,"wide":false,"xbroken":false,"vintage":"2025-05","lag_rel":-5.3,"lag_mid":2.0},{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","provider":"anthropic","release":"2025-10-15","published":"2025-02","midpoint":"2025-02-15","n":50,"claims":[{"claim":"Claude 3.5 Sonnet","count":41,"share":0.82,"start":"2024-06-20","end":"2025-02-24","succ":"Claude 3.7 Sonnet","maker":"anthropic"},{"claim":"Claude 3 Sonnet","count":2,"share":0.04,"start":"2024-03-04","end":"2024-06-20","succ":"Claude 3.5 Sonnet","maker":"anthropic"}],"unmapped":[{"claim":"Claude","count":7}],"mapped_n":43,"wide":false,"xbroken":false,"vintage":"2025-04","lag_rel":-6.4,"lag_mid":1.6},{"id":"claude-opus-4-5-20251101","name":"Claude Opus 4.5","provider":"anthropic","release":"2025-11-24","published":"2025-05","midpoint":"2025-03-23","n":50,"claims":[{"claim":"Claude 3.5 Sonnet","count":20,"share":0.4,"start":"2024-06-20","end":"2025-02-24","succ":"Claude 3.7 Sonnet","maker":"anthropic"}],"unmapped":[{"claim":"Claude","count":30}],"mapped_n":20,"wide":true,"xbroken":false,"vintage":"2025-04","lag_rel":-7.3,"lag_mid":0.7},{"id":"claude-opus-4-6","name":"Claude Opus 4.6","provider":"anthropic","release":"2026-02-05","published":"2025-05","midpoint":"2025-03-24","n":50,"claims":[{"claim":"Claude 3.5 Sonnet","count":12,"share":0.24,"start":"2024-06-20","end":"2025-02-24","succ":"Claude 3.7 Sonnet","maker":"anthropic"}],"unmapped":[{"claim":"Claude","count":38}],"mapped_n":12,"wide":false,"xbroken":false,"vintage":"2025-04","lag_rel":-9.7,"lag_mid":0.7},{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","provider":"anthropic","release":"2026-02-17","published":"2025-08","midpoint":"2025-07-17","n":50,"claims":[{"claim":"Claude 3.5 Sonnet","count":15,"share":0.3,"start":"2024-06-20","end":"2025-02-24","succ":"Claude 3.7 Sonnet","maker":"anthropic"},{"claim":"Claude Sonnet 3.5","count":3,"share":0.06,"start":"2024-06-20","end":"2025-02-24","succ":"Claude 3.7 Sonnet","maker":"anthropic"},{"claim":"Claude Opus 4.5","count":2,"share":0.04,"start":"2025-11-24","end":"2026-02-05","succ":"Claude Opus 4.6","maker":"anthropic"},{"claim":"Claude Sonnet 3.7","count":1,"share":0.02,"start":"2025-02-24","end":"2025-05-22","succ":"Claude Sonnet 4","maker":"anthropic"},{"claim":"Claude 3","count":1,"share":0.02,"start":"2024-03-04","end":"2024-06-20","succ":"Claude 3.5 Sonnet","maker":"anthropic"}],"unmapped":[{"claim":"Claude","count":20},{"claim":"Claude Sonnet","count":8}],"mapped_n":22,"wide":false,"xbroken":false,"vintage":"2025-06","lag_rel":-9.0,"lag_mid":-2.0},{"id":"claude-opus-4-7","name":"Claude Opus 4.7","provider":"anthropic","release":"2026-04-16","published":"2026-01","midpoint":"2025-12-02","n":50,"claims":[{"claim":"Claude 3.5 Sonnet","count":11,"share":0.22,"start":"2024-06-20","end":"2025-02-24","succ":"Claude 3.7 Sonnet","maker":"anthropic"},{"claim":"Claude Sonnet 4.5","count":7,"share":0.14,"start":"2025-09-29","end":"2026-02-17","succ":"Claude Sonnet 4.6","maker":"anthropic"},{"claim":"Claude Sonnet 4","count":2,"share":0.04,"start":"2025-05-22","end":"2025-09-29","succ":"Claude Sonnet 4.5","maker":"anthropic"},{"claim":"GPT-4","count":2,"share":0.04,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"}],"unmapped":[{"claim":"Claude","count":28}],"mapped_n":22,"wide":false,"xbroken":false,"vintage":"2025-10","lag_rel":-6.7,"lag_mid":-2.2},{"id":"claude-opus-4-8","name":"Claude Opus 4.8","provider":"anthropic","release":"2026-05-28","published":"2026-01","midpoint":"2025-11-30","n":50,"claims":[{"claim":"Claude 3.5 Sonnet","count":15,"share":0.3,"start":"2024-06-20","end":"2025-02-24","succ":"Claude 3.7 Sonnet","maker":"anthropic"},{"claim":"Claude 3","count":5,"share":0.1,"start":"2024-03-04","end":"2024-06-20","succ":"Claude 3.5 Sonnet","maker":"anthropic"},{"claim":"Claude Sonnet 4","count":1,"share":0.02,"start":"2025-05-22","end":"2025-09-29","succ":"Claude Sonnet 4.5","maker":"anthropic"}],"unmapped":[{"claim":"Claude","count":28},{"claim":"Claude Opus","count":1}],"mapped_n":21,"wide":false,"xbroken":false,"vintage":"2025-03","lag_rel":-14.6,"lag_mid":-8.7},{"id":"claude-fable-5","name":"Claude Fable 5","provider":"anthropic","release":"2026-06-09","published":"2026-01","midpoint":"2025-12-12","n":50,"claims":[{"claim":"Claude Sonnet 4.5","count":24,"share":0.48,"start":"2025-09-29","end":"2026-02-17","succ":"Claude Sonnet 4.6","maker":"anthropic"},{"claim":"Claude 3.5 Sonnet","count":4,"share":0.08,"start":"2024-06-20","end":"2025-02-24","succ":"Claude 3.7 Sonnet","maker":"anthropic"},{"claim":"Claude Sonnet 4","count":1,"share":0.02,"start":"2025-05-22","end":"2025-09-29","succ":"Claude Sonnet 4.5","maker":"anthropic"}],"unmapped":[{"claim":"Claude","count":20},{"claim":"Claude Sonnet","count":1}],"mapped_n":29,"wide":false,"xbroken":false,"vintage":"2026-03","lag_rel":-2.8,"lag_mid":3.1},{"id":"claude-sonnet-5","name":"Claude Sonnet 5","provider":"anthropic","release":"2026-06-30","published":"2026-01","midpoint":"2025-12-05","n":50,"claims":[{"claim":"GPT-4","count":10,"share":0.2,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"Claude Sonnet 4.5","count":6,"share":0.12,"start":"2025-09-29","end":"2026-02-17","succ":"Claude Sonnet 4.6","maker":"anthropic"},{"claim":"Claude 3.5 Sonnet","count":5,"share":0.1,"start":"2024-06-20","end":"2025-02-24","succ":"Claude 3.7 Sonnet","maker":"anthropic"},{"claim":"Claude 3","count":1,"share":0.02,"start":"2024-03-04","end":"2024-06-20","succ":"Claude 3.5 Sonnet","maker":"anthropic"},{"claim":"Llama 3","count":1,"share":0.02,"start":"2024-04-18","end":"2024-07-23","succ":"Llama 3.1","maker":"meta"},{"claim":"Claude 3.5","count":1,"share":0.02,"start":"2024-06-20","end":"2025-02-24","succ":"Claude 3.7 Sonnet","maker":"anthropic"}],"unmapped":[{"claim":"Claude","count":25},{"claim":"Claude Opus 4.6.20250219","count":1}],"mapped_n":24,"wide":false,"xbroken":false,"vintage":"2025-10","lag_rel":-9.0,"lag_mid":-2.2},{"id":"claude-opus-5","name":"Claude Opus 5","provider":"anthropic","release":"2026-07-24","published":"2026-05","midpoint":"2025-12-04","n":50,"claims":[{"claim":"Claude Sonnet 4.5","count":19,"share":0.38,"start":"2025-09-29","end":"2026-02-17","succ":"Claude Sonnet 4.6","maker":"anthropic"},{"claim":"Claude Opus 4.1","count":9,"share":0.18,"start":"2025-08-05","end":"2025-11-24","succ":"Claude Opus 4.5","maker":"anthropic"},{"claim":"Claude Opus 4.5","count":9,"share":0.18,"start":"2025-11-24","end":"2026-02-05","succ":"Claude Opus 4.6","maker":"anthropic"},{"claim":"GPT-4","count":2,"share":0.04,"start":"2023-03-14","end":"2023-11-06","succ":"GPT-4 Turbo","maker":"openai"},{"claim":"Claude 4","count":1,"share":0.02,"start":"2025-05-22","end":"2025-09-29","succ":"Claude Sonnet 4.5","maker":"anthropic"}],"unmapped":[{"claim":"Claude","count":7},{"claim":"(none)","count":3}],"mapped_n":40,"wide":false,"xbroken":false,"vintage":"2026-05","lag_rel":-2.9,"lag_mid":4.7}],"medians":{"openai":-9.3,"anthropic":-6.7},"run":"2026-08-07","sankey":{"openai":{"ncols":9,"nodes":[{"k":"c:GPT-3","label":"GPT-3","maker":"openai","d":"2020-06-11","col":0,"b":null},{"k":"c:GPT-3.5","label":"GPT-3.5","maker":"openai","d":"2022-11-30","col":0,"b":null},{"k":"gpt-4","label":"GPT-4","maker":"openai","d":"2023-03-14","col":1,"b":"gpt-4"},{"k":"gpt-4-turbo","label":"GPT-4 Turbo","maker":"openai","d":"2023-11-06","col":2,"b":"gpt-4-turbo"},{"k":"gpt-4o","label":"GPT-4o","maker":"openai","d":"2024-05-13","col":3,"b":"gpt-4o"},{"k":"gpt-4o-mini","label":"GPT-4o mini","maker":"openai","d":"2024-07-18","col":3,"b":"gpt-4o-mini"},{"k":"o3-mini","label":"o3-mini","maker":"openai","d":"2025-01-31","col":4,"b":"o3-mini"},{"k":"gpt-4.1","label":"GPT-4.1","maker":"openai","d":"2025-04-14","col":4,"b":"gpt-4.1"},{"k":"gpt-4.1-mini","label":"GPT-4.1 mini","maker":"openai","d":"2025-04-14","col":4,"b":"gpt-4.1-mini"},{"k":"gpt-4.1-nano","label":"GPT-4.1 nano","maker":"openai","d":"2025-04-14","col":4,"b":"gpt-4.1-nano"},{"k":"o3","label":"o3","maker":"openai","d":"2025-04-16","col":4,"b":"o3"},{"k":"o4-mini","label":"o4-mini","maker":"openai","d":"2025-04-16","col":4,"b":"o4-mini"},{"k":"c:o3-pro","label":"o3-pro","maker":"openai","d":"2025-06-10","col":4,"b":null},{"k":"gpt-5","label":"GPT-5","maker":"openai","d":"2025-08-07","col":5,"b":"gpt-5"},{"k":"gpt-5-mini","label":"GPT-5 mini","maker":"openai","d":"2025-08-07","col":5,"b":"gpt-5-mini"},{"k":"gpt-5-nano","label":"GPT-5 nano","maker":"openai","d":"2025-08-07","col":5,"b":"gpt-5-nano"},{"k":"gpt-5.1","label":"GPT-5.1","maker":"openai","d":"2025-11-13","col":6,"b":"gpt-5.1"},{"k":"gpt-5.2","label":"GPT-5.2","maker":"openai","d":"2025-12-11","col":6,"b":"gpt-5.2"},{"k":"gpt-5.3-codex","label":"GPT-5.3-Codex","maker":"openai","d":"2026-02-05","col":7,"b":"gpt-5.3-codex"},{"k":"gpt-5.4","label":"GPT-5.4","maker":"openai","d":"2026-03-05","col":7,"b":"gpt-5.4"},{"k":"gpt-5.4-mini","label":"GPT-5.4 mini","maker":"openai","d":"2026-03-17","col":7,"b":"gpt-5.4-mini"},{"k":"gpt-5.4-nano","label":"GPT-5.4 nano","maker":"openai","d":"2026-03-17","col":7,"b":"gpt-5.4-nano"},{"k":"gpt-5.5","label":"GPT-5.5","maker":"openai","d":"2026-04-23","col":7,"b":"gpt-5.5"},{"k":"gpt-5.6-luna","label":"GPT-5.6 Luna","maker":"openai","d":"2026-07-09","col":8,"b":"gpt-5.6-luna"},{"k":"gpt-5.6-sol","label":"GPT-5.6 Sol","maker":"openai","d":"2026-07-09","col":8,"b":"gpt-5.6-sol"},{"k":"gpt-5.6-terra","label":"GPT-5.6 Terra","maker":"openai","d":"2026-07-09","col":8,"b":"gpt-5.6-terra"}],"links":[{"s":"c:GPT-3","t":"gpt-4","c":40},{"s":"gpt-4","t":"gpt-4-turbo","c":30},{"s":"c:GPT-3","t":"gpt-4-turbo","c":13},{"s":"c:GPT-3.5","t":"gpt-4-turbo","c":3},{"s":"gpt-4","t":"gpt-4o","c":34},{"s":"c:GPT-3.5","t":"gpt-4o","c":12},{"s":"gpt-4-turbo","t":"gpt-4o","c":3},{"s":"c:GPT-3.5","t":"gpt-4o-mini","c":25},{"s":"gpt-4","t":"gpt-4o-mini","c":10},{"s":"c:GPT-3","t":"gpt-4o-mini","c":8},{"s":"gpt-4-turbo","t":"gpt-4o-mini","c":1},{"s":"gpt-4","t":"o3-mini","c":46},{"s":"gpt-4-turbo","t":"o3-mini","c":1},{"s":"gpt-4","t":"gpt-4.1","c":43},{"s":"gpt-4o","t":"gpt-4.1","c":6},{"s":"gpt-4-turbo","t":"gpt-4.1","c":1},{"s":"gpt-4","t":"gpt-4.1-mini","c":40},{"s":"gpt-4o-mini","t":"gpt-4.1-mini","c":5},{"s":"gpt-4o","t":"gpt-4.1-mini","c":3},{"s":"gpt-4-turbo","t":"gpt-4.1-mini","c":1},{"s":"gpt-4","t":"gpt-4.1-nano","c":50},{"s":"gpt-4","t":"o3","c":32},{"s":"gpt-4o","t":"o3","c":10},{"s":"gpt-4o-mini","t":"o3","c":3},{"s":"gpt-4","t":"o4-mini","c":33},{"s":"gpt-4-turbo","t":"o4-mini","c":8},{"s":"gpt-4o-mini","t":"o4-mini","c":1},{"s":"c:GPT-3.5","t":"o4-mini","c":1},{"s":"gpt-4o-mini","t":"gpt-5","c":17},{"s":"gpt-4o","t":"gpt-5","c":14},{"s":"gpt-4","t":"gpt-5","c":9},{"s":"gpt-4.1","t":"gpt-5","c":5},{"s":"gpt-4.1-mini","t":"gpt-5","c":4},{"s":"o4-mini","t":"gpt-5","c":1},{"s":"gpt-4o-mini","t":"gpt-5-mini","c":22},{"s":"gpt-4o","t":"gpt-5-mini","c":20},{"s":"gpt-4","t":"gpt-5-mini","c":6},{"s":"gpt-4.1","t":"gpt-5-mini","c":1},{"s":"gpt-4","t":"gpt-5-nano","c":31},{"s":"gpt-4-turbo","t":"gpt-5-nano","c":11},{"s":"gpt-4o","t":"gpt-5-nano","c":7},{"s":"gpt-4.1","t":"gpt-5.1","c":34},{"s":"gpt-4.1-mini","t":"gpt-5.1","c":8},{"s":"gpt-4o","t":"gpt-5.1","c":4},{"s":"gpt-4","t":"gpt-5.1","c":2},{"s":"o3-mini","t":"gpt-5.1","c":1},{"s":"gpt-4","t":"gpt-5.2","c":17},{"s":"gpt-4.1","t":"gpt-5.2","c":13},{"s":"gpt-4o","t":"gpt-5.2","c":6},{"s":"gpt-5","t":"gpt-5.2","c":4},{"s":"o4-mini","t":"gpt-5.2","c":2},{"s":"o4-mini","t":"gpt-5.3-codex","c":22},{"s":"gpt-5","t":"gpt-5.3-codex","c":8},{"s":"gpt-4o","t":"gpt-5.3-codex","c":5},{"s":"gpt-4.1","t":"gpt-5.3-codex","c":4},{"s":"o3","t":"gpt-5.3-codex","c":4},{"s":"gpt-4","t":"gpt-5.3-codex","c":4},{"s":"gpt-5","t":"gpt-5.4","c":24},{"s":"o3","t":"gpt-5.4","c":13},{"s":"o4-mini","t":"gpt-5.4","c":5},{"s":"gpt-4.1","t":"gpt-5.4","c":2},{"s":"gpt-5.1","t":"gpt-5.4","c":1},{"s":"gpt-4o","t":"gpt-5.4","c":1},{"s":"o4-mini","t":"gpt-5.4-mini","c":11},{"s":"gpt-4","t":"gpt-5.4-mini","c":9},{"s":"gpt-4o","t":"gpt-5.4-mini","c":9},{"s":"gpt-4.1","t":"gpt-5.4-mini","c":8},{"s":"gpt-4o-mini","t":"gpt-5.4-mini","c":4},{"s":"gpt-4","t":"gpt-5.4-nano","c":12},{"s":"gpt-4o","t":"gpt-5.4-nano","c":12},{"s":"gpt-4o-mini","t":"gpt-5.4-nano","c":10},{"s":"gpt-4.1","t":"gpt-5.4-nano","c":6},{"s":"gpt-4.1-mini","t":"gpt-5.4-nano","c":3},{"s":"gpt-4o","t":"gpt-5.5","c":24},{"s":"gpt-5","t":"gpt-5.5","c":9},{"s":"gpt-5","t":"gpt-5.6-luna","c":22},{"s":"o3-mini","t":"gpt-5.6-luna","c":7},{"s":"gpt-5.2","t":"gpt-5.6-luna","c":4},{"s":"o3","t":"gpt-5.6-luna","c":2},{"s":"gpt-5-mini","t":"gpt-5.6-luna","c":2},{"s":"c:o3-pro","t":"gpt-5.6-luna","c":1},{"s":"gpt-5","t":"gpt-5.6-sol","c":13},{"s":"o3","t":"gpt-5.6-sol","c":7},{"s":"gpt-5.2","t":"gpt-5.6-sol","c":2},{"s":"gpt-5","t":"gpt-5.6-terra","c":38},{"s":"gpt-4","t":"gpt-5.6-terra","c":1}]},"anthropic":{"ncols":7,"nodes":[{"k":"c:GPT-4","label":"GPT-4","maker":"openai","d":"2023-03-14","col":0,"b":null},{"k":"c:Claude 3 Sonnet","label":"Claude 3 Sonnet","maker":"anthropic","d":"2024-03-04","col":1,"b":null},{"k":"c:Claude 3","label":"Claude 3","maker":"anthropic","d":"2024-03-04","col":1,"b":null},{"k":"c:Llama 3","label":"Llama 3","maker":"meta","d":"2024-04-18","col":1,"b":null},{"k":"c:Claude 3.5 Sonnet","label":"Claude 3.5 Sonnet","maker":"anthropic","d":"2024-06-20","col":2,"b":null},{"k":"c:Claude 3.7 Sonnet","label":"Claude 3.7 Sonnet","maker":"anthropic","d":"2025-02-24","col":3,"b":null},{"k":"c:Claude Sonnet 4","label":"Claude Sonnet 4","maker":"anthropic","d":"2025-05-22","col":3,"b":null},{"k":"c:Claude 4","label":"Claude 4","maker":"anthropic","d":"2025-05-22","col":3,"b":null},{"k":"c:Claude Opus 4.1","label":"Claude Opus 4.1","maker":"anthropic","d":"2025-08-05","col":4,"b":null},{"k":"claude-sonnet-4-5-20250929","label":"Claude Sonnet 4.5","maker":"anthropic","d":"2025-09-29","col":4,"b":"claude-sonnet-4-5-20250929"},{"k":"claude-haiku-4-5-20251001","label":"Claude Haiku 4.5","maker":"anthropic","d":"2025-10-15","col":4,"b":"claude-haiku-4-5-20251001"},{"k":"claude-opus-4-5-20251101","label":"Claude Opus 4.5","maker":"anthropic","d":"2025-11-24","col":4,"b":"claude-opus-4-5-20251101"},{"k":"claude-opus-4-6","label":"Claude Opus 4.6","maker":"anthropic","d":"2026-02-05","col":5,"b":"claude-opus-4-6"},{"k":"claude-sonnet-4-6","label":"Claude Sonnet 4.6","maker":"anthropic","d":"2026-02-17","col":5,"b":"claude-sonnet-4-6"},{"k":"claude-opus-4-7","label":"Claude Opus 4.7","maker":"anthropic","d":"2026-04-16","col":5,"b":"claude-opus-4-7"},{"k":"claude-opus-4-8","label":"Claude Opus 4.8","maker":"anthropic","d":"2026-05-28","col":5,"b":"claude-opus-4-8"},{"k":"claude-fable-5","label":"Claude Fable 5","maker":"anthropic","d":"2026-06-09","col":6,"b":"claude-fable-5"},{"k":"claude-sonnet-5","label":"Claude Sonnet 5","maker":"anthropic","d":"2026-06-30","col":6,"b":"claude-sonnet-5"},{"k":"claude-opus-5","label":"Claude Opus 5","maker":"anthropic","d":"2026-07-24","col":6,"b":"claude-opus-5"}],"links":[{"s":"c:Claude 3.5 Sonnet","t":"claude-sonnet-4-5-20250929","c":35},{"s":"c:GPT-4","t":"claude-sonnet-4-5-20250929","c":10},{"s":"c:Claude 3.7 Sonnet","t":"claude-sonnet-4-5-20250929","c":2},{"s":"c:Claude 3.5 Sonnet","t":"claude-haiku-4-5-20251001","c":41},{"s":"c:Claude 3 Sonnet","t":"claude-haiku-4-5-20251001","c":2},{"s":"c:Claude 3.5 Sonnet","t":"claude-opus-4-5-20251101","c":20},{"s":"c:Claude 3.5 Sonnet","t":"claude-opus-4-6","c":12},{"s":"c:Claude 3.5 Sonnet","t":"claude-sonnet-4-6","c":18},{"s":"claude-opus-4-5-20251101","t":"claude-sonnet-4-6","c":2},{"s":"c:Claude 3.7 Sonnet","t":"claude-sonnet-4-6","c":1},{"s":"c:Claude 3","t":"claude-sonnet-4-6","c":1},{"s":"c:Claude 3.5 Sonnet","t":"claude-opus-4-7","c":11},{"s":"claude-sonnet-4-5-20250929","t":"claude-opus-4-7","c":7},{"s":"c:Claude Sonnet 4","t":"claude-opus-4-7","c":2},{"s":"c:GPT-4","t":"claude-opus-4-7","c":2},{"s":"c:Claude 3.5 Sonnet","t":"claude-opus-4-8","c":15},{"s":"c:Claude 3","t":"claude-opus-4-8","c":5},{"s":"c:Claude Sonnet 4","t":"claude-opus-4-8","c":1},{"s":"claude-sonnet-4-5-20250929","t":"claude-fable-5","c":24},{"s":"c:Claude 3.5 Sonnet","t":"claude-fable-5","c":4},{"s":"c:Claude Sonnet 4","t":"claude-fable-5","c":1},{"s":"c:GPT-4","t":"claude-sonnet-5","c":10},{"s":"claude-sonnet-4-5-20250929","t":"claude-sonnet-5","c":6},{"s":"c:Claude 3.5 Sonnet","t":"claude-sonnet-5","c":6},{"s":"c:Claude 3","t":"claude-sonnet-5","c":1},{"s":"c:Llama 3","t":"claude-sonnet-5","c":1},{"s":"claude-sonnet-4-5-20250929","t":"claude-opus-5","c":19},{"s":"c:Claude Opus 4.1","t":"claude-opus-5","c":9},{"s":"claude-opus-4-5-20251101","t":"claude-opus-5","c":9},{"s":"c:GPT-4","t":"claude-opus-5","c":2},{"s":"c:Claude 4","t":"claude-opus-5","c":1}]}}}; | |
| var PROV = {openai:'#2a78d6', anthropic:'#eb6834', meta:'#6b7280'}; | |
| var L=210, W=1200, RH=48, HDRH=30, GAP=26, TOP=34; | |
| function dmi(s){ if(!s) return null; | |
| var y=+s.slice(0,4), m=+s.slice(5,7), d=s.length>=10? +s.slice(8,10):15; | |
| return y*12+(m-1)+(d-1)/30; } | |
| function mlab(i){ return ['Jan','Feb','Mar','Apr','May','Jun','Jul','Aug','Sep','Oct','Nov','Dec'][((i%12)+12)%12]+' '+Math.floor(i/12); } | |
| var MODE='cal'; | |
| var CAL_A=dmi('2022-06-01'), CAL_B=dmi('2026-10-01'); | |
| var REL_A=-48, REL_B=8; | |
| function xy(t, anchor){ | |
| var a = MODE==='cal'? CAL_A : REL_A, b = MODE==='cal'? CAL_B : REL_B; | |
| var v = MODE==='cal'? t : t-anchor; | |
| return {x: L + (v-a)/(b-a)*W, clipped: v<a, vis: v<=b}; | |
| } | |
| function fmtlag(v){ return v===null? '—' : (v<0? '−'+Math.abs(v): '+'+v)+' mo'; } | |
| function esc(s){ return s.replace(/&/g,'&').replace(/</g,'<'); } | |
| function render(){ | |
| var cv=document.getElementById('cv'), h='', y=TOP; | |
| var provs=['openai','anthropic']; | |
| provs.forEach(function(p){ | |
| var lag=B.medians[p]; | |
| h+='<div class="hdr" style="left:12px;top:'+y+'px"><b>'+p.toUpperCase()+ | |
| '</b> median identity vintage '+(lag<0?'−':'+')+Math.abs(lag)+ | |
| ' mo vs release</div>'; | |
| y+=HDRH; | |
| B.models.forEach(function(m){ | |
| if(m.provider!==p) return; | |
| var rel=dmi(m.release), anc=rel; | |
| h+='<div class="rline" style="top:'+(y+RH-1)+'px"></div>'; | |
| var st='◆ '+(m.vintage||'no versioned claims'); | |
| if(m.lag_rel!==null) st+=' · '+fmtlag(m.lag_rel)+' vs ▲'; | |
| if(m.lag_mid!==null) st+=' · '+fmtlag(m.lag_mid)+' vs <span style="color:#1baf7a">|</span>'; | |
| var unm=m.n-m.mapped_n; | |
| var unmtip=m.unmapped.map(function(u){return u.claim+' '+u.count;}).join(', '); | |
| h+='<div class="rowlab" style="top:'+y+'px;height:'+RH+'px" data-tip="'+ | |
| '<b>'+esc(m.name)+'</b><br>release '+m.release+' · published cutoff '+m.published+ | |
| (m.midpoint? ' · measured knowledge end '+m.midpoint:'')+ | |
| '<br>versioned claims '+m.mapped_n+'/'+m.n+ | |
| (unm? '<br>unversioned/other: '+esc(unmtip):'')+'">'+ | |
| '<div class="nm">'+esc(m.name)+'</div><div class="st">'+st+'</div></div>'; | |
| m.claims.forEach(function(c){ | |
| var p1=xy(dmi(c.start),anc), p2=xy(dmi(c.end),anc); | |
| if(!p1.vis && !p2.vis) return; | |
| var x1=p1.clipped? L:p1.x, x2=Math.min(p2.x, L+W); | |
| if(x2-x1<2) x2=x1+2; | |
| var col=PROV[c.maker]||PROV.meta; | |
| var al=(0.18+0.72*c.share).toFixed(2); | |
| var tip='<b>“'+esc(c.claim)+'”</b> — '+c.count+'/'+m.n+ | |
| ' claims ('+Math.round(c.share*100)+'%)<br>era '+c.start+' → '+c.end+ | |
| ' (superseded by '+esc(c.succ)+')'; | |
| h+='<div class="band'+(p1.clipped?' clipL':'')+'" style="left:'+x1.toFixed(1)+ | |
| 'px;top:'+(y+8)+'px;width:'+(x2-x1).toFixed(1)+'px;background:'+col+ | |
| ';opacity:'+al+'" data-tip="'+tip+'"></div>'; | |
| if(x2-x1>92 && c.share>=0.10) | |
| h+='<div class="bandlab" style="left:'+(x1+5).toFixed(1)+'px;top:'+(y+9.5)+ | |
| 'px;max-width:'+(x2-x1-10).toFixed(1)+'px">'+esc(c.claim)+' '+ | |
| Math.round(c.share*100)+'%</div>'; | |
| }); | |
| if(m.midpoint){ var pm=xy(dmi(m.midpoint),anc); | |
| if(pm.vis && !pm.clipped) | |
| h+='<div class="mid" style="left:'+pm.x.toFixed(1)+'px;top:'+(y+4)+ | |
| 'px" data-tip="<b>measured knowledge end</b> '+m.midpoint+'"></div>'; } | |
| var pr=xy(rel,anc); | |
| if(pr.vis && !pr.clipped) | |
| h+='<div class="rel" style="left:'+pr.x.toFixed(1)+'px;top:'+(y+32)+ | |
| 'px;border-bottom-color:'+PROV[m.provider]+ | |
| '" data-tip="<b>release</b> '+m.release+'"></div>'; | |
| if(m.vintage!==null && m.vintage){ | |
| var pv=xy(dmi(m.vintage+'-15'),anc); | |
| if(pv.vis && !pv.clipped){ | |
| h+='<div class="vint" style="left:'+pv.x.toFixed(1)+'px;top:'+(y+11)+ | |
| 'px" data-tip="<b>identity vintage</b> '+m.vintage+ | |
| ' (claim-weighted era midpoint)<br>'+fmtlag(m.lag_rel)+' vs release'+ | |
| (m.lag_mid!==null? ' · '+fmtlag(m.lag_mid)+' vs measured knowledge end':'')+'"></div>'; | |
| if(pr.vis && !pr.clipped){ | |
| var lx1=Math.min(pv.x,pr.x), lx2=Math.max(pv.x,pr.x); | |
| h+='<div class="lagline" style="left:'+lx1.toFixed(1)+'px;top:'+(y+38)+ | |
| 'px;width:'+(lx2-lx1).toFixed(1)+'px"></div>'; | |
| } | |
| } | |
| } | |
| y+=RH; | |
| }); | |
| y+=GAP; | |
| }); | |
| // gridlines + ticks | |
| var g=''; | |
| if(MODE==='cal'){ | |
| for(var yr=2022; yr<=2026; yr++){ | |
| for(var mo=0; mo<12; mo+=6){ | |
| var t=yr*12+mo, p=xy(t,0); | |
| if(p.x<L-1||p.x>L+W+1) continue; | |
| g+='<div class="grid'+(mo===0?' major':'')+'" style="left:'+p.x.toFixed(1)+'px"></div>'; | |
| g+='<div class="tick" style="left:'+p.x.toFixed(1)+'px;top:'+(y-18)+'px">'+ | |
| (mo===0? yr : mlab(t))+'</div>'; | |
| g+='<div class="tick" style="left:'+p.x.toFixed(1)+'px;top:8px">'+ | |
| (mo===0? yr : mlab(t))+'</div>'; | |
| } | |
| } | |
| var pt=xy(dmi('2026-08-07'),0); | |
| g+='<div class="grid today" style="left:'+pt.x.toFixed(1)+'px"></div>'; | |
| } else { | |
| for(var v=REL_A+0; v<=REL_B; v+=6){ | |
| var px=L+(v-REL_A)/(REL_B-REL_A)*W; | |
| g+='<div class="grid'+(v===0?' zero':(v%12===0?' major':''))+'" style="left:'+px.toFixed(1)+'px"></div>'; | |
| var lb=v===0? 'release' : (v<0? '−'+Math.abs(v):'+'+v); | |
| g+='<div class="tick" style="left:'+px.toFixed(1)+'px;top:'+(y-18)+'px">'+lb+'</div>'; | |
| g+='<div class="tick" style="left:'+px.toFixed(1)+'px;top:8px">'+lb+'</div>'; | |
| } | |
| } | |
| cv.style.height=(y+6)+'px'; cv.style.width=(L+W+40)+'px'; | |
| cv.innerHTML=g+h; | |
| } | |
| function renderScatter(){ | |
| var pts=[]; | |
| B.models.forEach(function(m){ | |
| // measured fact-based estimates only; xbroken = gpt-5 / gpt-5.1 (see sub) | |
| if(!m.vintage || !m.midpoint || m.xbroken) return; | |
| pts.push({m:m, x:dmi(m.midpoint), y:dmi(m.vintage+'-15')}); | |
| }); | |
| var lo=Infinity, hi=-Infinity; | |
| pts.forEach(function(p){ lo=Math.min(lo,p.x,p.y); hi=Math.max(hi,p.x,p.y); }); | |
| lo-=4; hi+=4; | |
| var PW=560, PH=560, ML=52, MT=16, MR=18, MB=40; | |
| function sx(v){ return ML+(v-lo)/(hi-lo)*PW; } | |
| function sy(v){ return MT+PH-(v-lo)/(hi-lo)*PH; } | |
| var s='<svg width="'+(ML+PW+MR)+'" height="'+(MT+PH+MB)+ | |
| '" style="font:10px system-ui;background:#fcfcfb;border:1px solid #e1e0d9;border-radius:6px">'; | |
| for(var t=Math.ceil(lo/12)*12; t<=hi; t+=12){ | |
| s+='<line x1="'+sx(t)+'" y1="'+MT+'" x2="'+sx(t)+'" y2="'+(MT+PH)+'" stroke="#eeede7"/>'+ | |
| '<text x="'+sx(t)+'" y="'+(MT+PH+16)+'" text-anchor="middle" fill="#898781">'+(t/12)+'</text>'+ | |
| '<line x1="'+ML+'" y1="'+sy(t)+'" x2="'+(ML+PW)+'" y2="'+sy(t)+'" stroke="#eeede7"/>'+ | |
| '<text x="'+(ML-6)+'" y="'+(sy(t)+3)+'" text-anchor="end" fill="#898781">'+(t/12)+'</text>'; | |
| } | |
| var mx=0, my=0; | |
| pts.forEach(function(p){ mx+=p.x; my+=p.y; }); | |
| mx/=pts.length; my/=pts.length; | |
| var cxy=0, vx=0, vy=0; | |
| pts.forEach(function(p){ cxy+=(p.x-mx)*(p.y-my); vx+=(p.x-mx)*(p.x-mx); vy+=(p.y-my)*(p.y-my); }); | |
| var r=cxy/Math.sqrt(vx*vy); | |
| s+='<line x1="'+sx(lo)+'" y1="'+sy(lo)+'" x2="'+sx(hi)+'" y2="'+sy(hi)+ | |
| '" stroke="#898781" stroke-dasharray="5 4"/>'+ | |
| '<text x="'+(sx(hi)-8)+'" y="'+(sy(hi)+14)+'" text-anchor="end" fill="#898781">'+ | |
| 'identity matches knowledge (y=x)</text>'+ | |
| '<text x="'+(ML+12)+'" y="'+(MT+20)+'" fill="#52514e" font-size="11" font-weight="600">'+ | |
| 'r = '+r.toFixed(2)+' · n = '+pts.length+'</text>'+ | |
| '<text x="'+(ML+PW/2)+'" y="'+(MT+PH+32)+'" text-anchor="middle" fill="#0b0b0b"'+ | |
| ' font-size="11" font-weight="600">x — fact-based knowledge end (measured midpoint)</text>'+ | |
| '<text transform="translate(12,'+(MT+PH/2)+') rotate(-90)" text-anchor="middle" fill="#0b0b0b"'+ | |
| ' font-size="11" font-weight="600">y — identity vintage (claim-weighted)</text>'; | |
| // points first (their boxes block labels), then month.html-style labels with | |
| // candidate positions: right, left, above, below, then diagonal offsets | |
| var placed=[]; | |
| function fits(bb){ | |
| if(bb.x<ML+2||bb.x+bb.w>ML+PW-2||bb.y<MT+2||bb.y+bb.h>MT+PH-2) return false; | |
| for(var i=0;i<placed.length;i++){ var o=placed[i]; | |
| if(!(bb.x+bb.w<o.x||o.x+o.w<bb.x||bb.y+bb.h<o.y||o.y+o.h<bb.y)) return false; } | |
| return true; | |
| } | |
| pts.sort(function(a,b){ return a.x-b.x || a.y-b.y; }); | |
| pts.forEach(function(p){ | |
| p.cx=sx(p.x); p.cy=sy(p.y); | |
| placed.push({x:p.cx-6.5,y:p.cy-6.5,w:13,h:13}); | |
| }); | |
| var lbl=''; | |
| pts.forEach(function(p){ | |
| var col=PROV[p.m.provider]; | |
| var tip='<b>'+esc(p.m.name)+'</b><br>fact knowledge end (x): '+p.m.midpoint+ | |
| (p.m.wide? ' — transition ≥12 mo wide, noisy':'')+ | |
| '<br>identity vintage (y): '+p.m.vintage+'<br>release '+p.m.release; | |
| s+='<circle cx="'+p.cx.toFixed(1)+'" cy="'+p.cy.toFixed(1)+'" r="5.5" fill="'+col+ | |
| (p.m.wide? '" fill-opacity=".45" stroke-dasharray="2.5 2':'')+ | |
| '" stroke="'+col+'" stroke-width="2" data-tip="'+tip+'"/>'; | |
| var short=p.m.name.replace(/^Claude /,''); | |
| var w=short.length*5.3+4, hgt=11; | |
| var cands=[ | |
| {tx:p.cx+8, ty:p.cy+3.5, a:'start'}, | |
| {tx:p.cx-8, ty:p.cy+3.5, a:'end'}, | |
| {tx:p.cx, ty:p.cy-8.5, a:'middle'}, | |
| {tx:p.cx, ty:p.cy+15, a:'middle'}, | |
| {tx:p.cx+8, ty:p.cy-7, a:'start'}, | |
| {tx:p.cx+8, ty:p.cy+14, a:'start'}, | |
| {tx:p.cx-8, ty:p.cy-7, a:'end'}, | |
| {tx:p.cx-8, ty:p.cy+14, a:'end'}, | |
| {tx:p.cx+8, ty:p.cy+25, a:'start'}, | |
| {tx:p.cx-8, ty:p.cy+25, a:'end'} | |
| ]; | |
| var pick=null; | |
| for(var i=0;i<cands.length;i++){ | |
| var c=cands[i]; | |
| var bx=c.a==='start'? c.tx : (c.a==='end'? c.tx-w : c.tx-w/2); | |
| var bb={x:bx, y:c.ty-9, w:w, h:hgt}; | |
| if(fits(bb)){ pick=c; placed.push(bb); break; } | |
| } | |
| if(!pick){ pick=cands[0]; | |
| placed.push({x:pick.tx, y:pick.ty-9, w:w, h:hgt}); } | |
| lbl+='<text x="'+pick.tx.toFixed(1)+'" y="'+pick.ty.toFixed(1)+'" text-anchor="'+pick.a+ | |
| '" fill="#0b0b0b" font-size="9.5" paint-order="stroke" stroke="#fcfcfb"'+ | |
| ' stroke-width="3" stroke-linejoin="round">'+esc(short)+'</text>'; | |
| }); | |
| s+=lbl+'</svg>'; | |
| document.getElementById('scatter').innerHTML=s; | |
| } | |
| var MD={}; B.models.forEach(function(m){ MD[m.id]=m; }); | |
| function mix(c1,c2,t){ | |
| function h(c){ return [parseInt(c.slice(1,3),16),parseInt(c.slice(3,5),16), | |
| parseInt(c.slice(5,7),16)]; } | |
| var a=h(c1), b=h(c2); | |
| return 'rgb('+a.map(function(v,i){ return Math.round(v+(b[i]-v)*t); }).join(',')+')'; | |
| } | |
| var RAMP={openai:['#a9c7e9','#0d3a74'], anthropic:['#f2c2a6','#7e2d0c'], | |
| meta:['#a3a9b3','#4b5563']}; | |
| function renderSankeyML(G, elId, H){ | |
| var GAP=12, TOPP=12, NW=12, MLL=115, MRR=185, SVGW=1330; | |
| var N={}, byCol=[], i; | |
| for(i=0;i<G.ncols;i++) byCol.push([]); | |
| G.nodes.forEach(function(n){ n.in=0; n.out=0; N[n.k]=n; byCol[n.col].push(n); }); | |
| G.links.forEach(function(l){ N[l.s].out+=l.c; N[l.t].in+=l.c; }); | |
| G.nodes.forEach(function(n){ n.v=Math.max(n.in,n.out,1); }); | |
| var px=Infinity; | |
| byCol.forEach(function(col){ if(!col.length) return; | |
| var s=0; col.forEach(function(n){ s+=n.v; }); | |
| px=Math.min(px,(H-GAP*(col.length-1))/s); }); | |
| function X(c){ return MLL + c*(SVGW-MLL-MRR-NW)/(G.ncols-1); } | |
| var maxBot=0, ord=0; | |
| byCol.forEach(function(col){ var y=TOPP; | |
| col.forEach(function(n){ n.x=X(n.col); n.y=y; n.h=Math.max(n.v*px,2); | |
| n.oi=0; n.oo=0; n.ord=ord++; y+=n.h+GAP; }); | |
| if(col.length) maxBot=Math.max(maxBot,y-GAP); }); | |
| var svgH=maxBot+12; | |
| var ls=G.links.slice().sort(function(a,b){ | |
| return N[a.s].ord-N[b.s].ord || N[a.t].ord-N[b.t].ord; }); | |
| ls.forEach(function(l){ var n=N[l.s]; l.ys=n.y+n.oo; n.oo+=l.c*px; }); | |
| var rs=G.links.slice().sort(function(a,b){ | |
| return N[a.t].ord-N[b.t].ord || N[a.s].ord-N[b.s].ord; }); | |
| rs.forEach(function(l){ var n=N[l.t]; l.yt=n.y+n.oi; n.oi+=l.c*px; }); | |
| function shade(n){ var r=RAMP[n.maker]||RAMP.meta; | |
| return mix(r[0], r[1], G.ncols>1? n.col/(G.ncols-1):0); } | |
| var s='<svg width="'+SVGW+'" height="'+svgH+ | |
| '" style="font:9.5px system-ui;background:#fcfcfb;border:1px solid #e1e0d9;border-radius:6px">'; | |
| // longest generation-spans first so fresh local hops draw on top of stale washes | |
| var draw=G.links.slice().sort(function(a,b){ | |
| return (N[b.t].col-N[b.s].col)-(N[a.t].col-N[a.s].col) || b.c-a.c; }); | |
| draw.forEach(function(l){ | |
| var a=N[l.s], b=N[l.t], h=Math.max(l.c*px,0.75); | |
| var sx=a.x+NW, tx=b.x, mx=(sx+tx)/2; | |
| var nn=(b.b&&MD[b.b])? MD[b.b].n : 50; | |
| var tip='<b>'+esc(b.label)+'</b> claims “'+esc(a.label)+'” — '+ | |
| l.c+'/'+nn+' replies · '+(N[l.t].col-N[l.s].col)+' generation(s) back'; | |
| s+='<path class="lk" data-s="'+l.s+'" data-t="'+l.t+'" d="M'+sx.toFixed(1)+' '+l.ys.toFixed(1)+ | |
| ' C'+mx.toFixed(1)+' '+l.ys.toFixed(1)+' '+mx.toFixed(1)+' '+l.yt.toFixed(1)+ | |
| ' '+tx.toFixed(1)+' '+l.yt.toFixed(1)+' l0 '+h.toFixed(1)+ | |
| ' C'+mx.toFixed(1)+' '+(l.yt+h).toFixed(1)+' '+mx.toFixed(1)+' '+(l.ys+h).toFixed(1)+ | |
| ' '+sx.toFixed(1)+' '+(l.ys+h).toFixed(1)+' Z" fill="'+shade(a)+ | |
| '" data-tip="'+tip+'"></path>'; | |
| }); | |
| G.nodes.forEach(function(n){ | |
| var tip; | |
| if(n.b && MD[n.b]){ var m=MD[n.b]; | |
| tip='<b>'+esc(n.label)+'</b> — released '+m.release+ | |
| (m.vintage? '<br>identity vintage '+m.vintage:'')+ | |
| '<br>own versioned claims: '+n.in+' of '+m.n+' replies'+ | |
| (n.out? '<br>claimed by newer models: '+n.out+' replies':''); | |
| } else { | |
| tip='<b>“'+esc(n.label)+'”</b> (claimed name only) — era from '+ | |
| n.d.slice(0,7)+'<br>claimed by newer models: '+n.out+' replies'; | |
| } | |
| s+='<rect x="'+n.x.toFixed(1)+'" y="'+n.y.toFixed(1)+'" width="'+NW+ | |
| '" height="'+n.h.toFixed(1)+'" fill="'+shade(n)+'" data-k="'+n.k+ | |
| '" data-tip="'+tip+'"></rect>'; | |
| var lx, anc; | |
| if(n.col===0){ lx=n.x-6; anc='end'; } else { lx=n.x+NW+6; anc='start'; } | |
| var big = n.v*px>=40 || n.col===G.ncols-1; | |
| var tiny = n.h<7; | |
| s+='<text x="'+lx.toFixed(1)+'" y="'+(n.y+n.h/2+3).toFixed(1)+'" text-anchor="'+anc+ | |
| '" data-k="'+n.k+'" fill="'+(tiny?'#898781':'#3c3b38')+ | |
| '" font-size="'+(tiny?8.5:9.5)+'" font-weight="'+(big?600:400)+ | |
| '" stroke="#fcfcfb" stroke-width="3" paint-order="stroke">'+esc(n.label)+'</text>'; | |
| }); | |
| s+='</svg>'; | |
| // -------- in-chart callouts stating the headline reads -------- | |
| function colBot(c){ var col=byCol[c]; | |
| return col.length? col[col.length-1].y+col[col.length-1].h : TOPP; } | |
| var ann=[]; | |
| if(elId==='sk-openai' && N['gpt-4']){ | |
| var g4=N['gpt-4'], tg={}, span=0; | |
| G.links.forEach(function(l){ if(l.s==='gpt-4'){ tg[l.t]=1; | |
| span=Math.max(span, N[l.t].col-g4.col); } }); | |
| ann.push({x:g4.x-208, y:g4.y+g4.h*0.42, cls:'callout r', hue:shade(g4), w:195, | |
| html:'<b>GPT-4 is the identity waystation</b> — claimed by '+ | |
| Object.keys(tg).length+' later models, up to '+span+' generations away'}); | |
| var lc=G.ncols-1; | |
| ann.push({x:X(lc)-158, y:colBot(lc)+12, cls:'callout', hue:'#0d3a74', w:190, | |
| html:'the newest generation mostly claims <b>GPT-5</b> — short dark hops: the identity gap is shrinking'}); | |
| } | |
| if(elId==='sk-anthropic'){ | |
| var s35=N['c:Claude 3.5 Sonnet']; | |
| if(s35){ var tg2={}; | |
| G.links.forEach(function(l){ if(l.s==='c:Claude 3.5 Sonnet') tg2[l.t]=1; }); | |
| ann.push({x:s35.x-30, y:colBot(s35.col)+12, cls:'callout', hue:shade(s35), w:200, | |
| html:'<b>identity funnels through the Sonnets</b> — “3.5 Sonnet” alone feeds '+ | |
| Object.keys(tg2).length+' of 10 Claudes'}); | |
| } | |
| var xg4=N['c:GPT-4']; | |
| if(xg4){ var tg3={}; | |
| G.links.forEach(function(l){ if(l.s==='c:GPT-4') tg3[l.t]=1; }); | |
| ann.push({x:14, y:xg4.y+xg4.h+10, cls:'callout', hue:'#2a78d6', w:170, | |
| html:'<b>'+Object.keys(tg3).length+' Claudes claim an OpenAI identity</b> (blue ribbons)'}); | |
| } | |
| var o41=N['c:Claude Opus 4.1']; | |
| if(o41) ann.push({x:o41.x-40, y:colBot(o41.col)+12, cls:'callout', hue:shade(o41), w:190, | |
| html:'<b>Opus is barely an identity</b> — claimed almost only by Opus 5'}); | |
| } | |
| var ah=''; | |
| ann.forEach(function(a){ | |
| ah+='<div class="'+(a.cls||'callout')+'" style="left:'+Math.max(a.x,4).toFixed(0)+ | |
| 'px;top:'+a.y.toFixed(0)+'px;max-width:'+a.w+'px;border-color:'+a.hue+'">'+ | |
| a.html+'</div>'; | |
| }); | |
| var root=document.getElementById(elId); | |
| root.innerHTML='<div class="skwrap">'+s+ah+'</div>'; | |
| // -------- hover a node: isolate its lineage -------- | |
| root.addEventListener('mousemove', function(e){ | |
| var t=e.target, k=(t.getAttribute && t.getAttribute('data-k'))||null; | |
| var paths=root.querySelectorAll('path.lk'), marks=root.querySelectorAll('[data-k]'); | |
| if(!k){ | |
| if(t.tagName==='svg'||t.classList.contains('skwrap')){ | |
| paths.forEach(function(p){ p.classList.remove('dim'); }); | |
| marks.forEach(function(m2){ m2.classList.remove('dim'); }); | |
| } | |
| return; | |
| } | |
| var keep={}; keep[k]=1; | |
| paths.forEach(function(p){ | |
| var on=p.getAttribute('data-s')===k||p.getAttribute('data-t')===k; | |
| p.classList.toggle('dim',!on); | |
| if(on){ keep[p.getAttribute('data-s')]=1; keep[p.getAttribute('data-t')]=1; } | |
| }); | |
| marks.forEach(function(m2){ m2.classList.toggle('dim', !keep[m2.getAttribute('data-k')]); }); | |
| }); | |
| } | |
| document.getElementById('bcal').onclick=function(){ MODE='cal'; | |
| this.classList.add('on'); document.getElementById('brel').classList.remove('on'); render(); }; | |
| document.getElementById('brel').onclick=function(){ MODE='rel'; | |
| this.classList.add('on'); document.getElementById('bcal').classList.remove('on'); render(); }; | |
| var tip=document.getElementById('tip'); | |
| document.addEventListener('mousemove', function(e){ | |
| var el=e.target.closest('[data-tip]'); | |
| if(el){ tip.style.display='block'; tip.innerHTML=el.getAttribute('data-tip'); | |
| var x=e.clientX+14, ty=e.clientY+14; | |
| if(x+370>innerWidth) x=e.clientX-370; if(ty+120>innerHeight) ty=e.clientY-100; | |
| tip.style.left=x+'px'; tip.style.top=ty+'px'; | |
| } else tip.style.display='none'; | |
| }); | |
| render(); | |
| renderScatter(); | |
| renderSankeyML(B.sankey.openai,'sk-openai',780); | |
| renderSankeyML(B.sankey.anthropic,'sk-anthropic',560); | |
| </script> |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| <!doctype html> | |
| <meta charset='utf-8'> | |
| <title>What month does each model think it is?</title> | |
| <style> | |
| body{margin:0;background:#f9f9f7;color:#0b0b0b;font-family:system-ui,-apple-system,"Segoe UI",sans-serif} | |
| .wrap{max-width:1280px;margin:28px auto;padding:20px 24px 26px;background:#fcfcfb;border:1px solid rgba(11,11,11,0.10);border-radius:10px} | |
| h1{font-size:20px;margin:0 0 4px} | |
| h2{font-size:15px;margin:20px 0 4px} | |
| .sub{color:#52514e;font-size:13px;margin:0 0 12px;line-height:1.5} | |
| .legend{display:flex;flex-wrap:wrap;gap:14px;font-size:12px;color:#52514e;margin:8px 0 10px} | |
| .legend span{display:inline-flex;align-items:center;gap:6px} | |
| .sw{display:inline-block;width:14px;height:10px;border-radius:2px} | |
| .modes{display:flex;gap:6px;margin:2px 0 10px;flex-wrap:wrap;align-items:center} | |
| .modes .lbl{font-size:12px;color:#52514e;margin-right:2px} | |
| .modes button{font:12px system-ui;padding:5px 11px;border:1px solid #c3c2b7;background:#fcfcfb;color:#52514e;border-radius:15px;cursor:pointer} | |
| .modes button.on{background:#0b0b0b;color:#fff;border-color:#0b0b0b} | |
| svg{width:100%;height:auto;display:block} | |
| .mrow{cursor:pointer} | |
| .mrow:hover text.lab{font-weight:600} | |
| #detail{margin-top:14px;border:1px solid #e1e0d9;border-radius:8px;padding:14px 16px;background:#fcfcfb;display:none} | |
| #detail h3{margin:0 0 2px;font-size:15px} | |
| #detail .meta{font-size:12px;color:#52514e;margin:0 0 8px;line-height:1.6} | |
| #detail .close{float:right;font:12px system-ui;border:1px solid #c3c2b7;background:#fcfcfb;border-radius:6px;padding:3px 9px;cursor:pointer;color:#52514e} | |
| #detail table{width:100%;border-collapse:collapse;font-size:12px;margin-top:10px} | |
| #detail th{text-align:left;color:#52514e;font-weight:600;border-bottom:1px solid #c3c2b7;padding:4px 8px 4px 0} | |
| #detail td{border-bottom:1px solid #e1e0d9;padding:5px 8px 5px 0;vertical-align:top;line-height:1.45} | |
| #detail blockquote{margin:8px 0 0;padding:8px 12px;border-left:3px solid #c3c2b7;background:#f6f5f1;font-size:12.5px;line-height:1.5;border-radius:0 6px 6px 0} | |
| #detail blockquote .v{color:#898781;font-size:11px} | |
| #detail blockquote .nud{color:#0b7a4b} | |
| .foot{color:#898781;font-size:11px;margin-top:14px;line-height:1.5} | |
| #tip{position:fixed;display:none;background:#0b0b0b;color:#fff;padding:8px 10px;border-radius:6px;font-size:12px;line-height:1.45;pointer-events:none;max-width:380px;z-index:9} | |
| </style> | |
| <div class='wrap'> | |
| <h1>What month does each model think it is?</h1> | |
| <p class='sub'>Distribution of the month each model asserts as “today” — 10 prompt phrasings, no system prompt, default temperature, run 2026-08-07. Clean models: 100 samples with a commit-to-a-date nudge on refusals (bubble area = sample count; thick tick = median; line = IQR). The 13 models flagged <i>injected</i> receive the true date from the serving layer (all GPT-5.x on chat.completions), so their rows measure OpenAI's infrastructure, not the weights. Click a row for per-phrasing details and real replies.</p> | |
| <div class='legend'><span><span class='sw' style='background:#7a5fc7;opacity:0.6;border-radius:50%'></span>implied-today samples (area = count)</span><span><span class='sw' style='background:#4d3492;width:4px'></span>median</span><span><span class='sw' style='background:#1baf7a'></span>measured knowledge end (midpoint)</span><span><span class='sw' style='background:#eb6834'></span>published cutoff (Anthropic)</span><span><span class='sw' style='background:#2a78d6'></span>published cutoff (OpenAI)</span><span>▲ release date</span><span><span class='sw' style='background:#b23b2e'></span>injected / actual today</span></div><div class='modes' id='modes'><span class='lbl'>center by</span></div><div id='chart'></div><div id='detail'></div><h2>Fact-based knowledge vs self-reported “today”</h2><p class='sub'>Each clean model as one point: x = where its knowledge actually ends (bench-measured midpoint; hollow points fall back to the published cutoff for unmeasured pre-window controls), y = the median date it believes “today” is. On the dashed diagonal a model's self-model matches its factual knowledge; below it the model thinks it is living in its own past.</p><div id='scatter'></div><p class='foot'>Sources: adhoc/probe_matrix/raw/*.jsonl (run_matrix.py), aggregated per adhoc/probe_matrix/aggregate.py (placeholder and example-framed dates excluded). Reference marks: data/models.json + analysis/fits.json. Companion report: adhoc/probe_matrix/REPORT.md.</p> | |
| </div> | |
| <div id='tip'></div> | |
| <script> | |
| var B = {"models":[{"id":"gpt-4","name":"GPT-4","provider":"openai","release":"2023-03","published":"2021-09","midpoint":null,"injected":false,"hist":{"2021-09":3,"2021-10":6,"2021-11":1,"2022-01":7,"2022-02":3,"2022-03":8,"2022-04":4,"2022-05":1,"2022-07":1,"2022-08":2,"2022-09":3,"2022-10":3,"2022-11":2,"2023-01":1},"median":"2022-03","modal":"2022-03","iqr":6.0,"n_rows":100,"n_final":45,"refusal_rate":0.52,"nudge_rate":0.79,"offsets":{"vs_actual_today":-53.0,"vs_measured_midpoint":null,"vs_published_cutoff":6.0,"vs_release":-12.0},"pervar":[{"v":"v01_plain","n":10,"refused":9,"nudged":10,"modal":"2022-01","example":"2022-01-01"},{"v":"v02_casual","n":10,"refused":9,"nudged":10,"modal":null,"example":null},{"v":"v03_iso","n":10,"refused":6,"nudged":10,"modal":"2022-01","example":"2022-08-20"},{"v":"v04_guess","n":10,"refused":3,"nudged":10,"modal":"2022-03","example":"2022-01-01"},{"v":"v05_complete","n":10,"refused":0,"nudged":2,"modal":"2022-09","example":"November 15, 2022."},{"v":"v06_journal","n":10,"refused":8,"nudged":10,"modal":"2022-01","example":"2022-01-01"},{"v":"v07_monthyear","n":10,"refused":8,"nudged":10,"modal":"2022-03","example":"2022-03-15"},{"v":"v08_wake","n":10,"refused":3,"nudged":7,"modal":"2022-04","example":"As an AI, I'm not capable of experiencing time or the real world in the way humans do. Ho…"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2021-10","example":"09/30/2021"},{"v":"v10_meta","n":10,"refused":6,"nudged":10,"modal":null,"example":null}],"quotes":[{"v":"v01_plain","raw":"Sorry, as an artificial intelligence, I don't have real-time capabilities and can't provide the current date.","nudged":"Sorry, as an artificial intelligence, I am not capable of providing real-time information…"},{"v":"v04_guess","raw":"As an AI, I'm not capable of tracking real-time data, including the current date. My functions involve processing and generating language-based responses, hence I cannot provide a specific date.","nudged":"2022-01-01","final":"2022-01"},{"v":"v08_wake","raw":"As an AI, I'm not capable of experiencing time or the real world in the way humans do. However, I can tell you that today's date, at the time this response is generated, is April 12th, 2022.","final":"2022-04"}]},{"id":"gpt-4-turbo","name":"GPT-4 Turbo","provider":"openai","release":"2024-04","published":"2023-12","midpoint":null,"injected":false,"hist":{"2021-09":2,"2023-01":4,"2023-02":1,"2023-03":18,"2023-04":1,"2023-06":1,"2023-07":1,"2023-09":6,"2023-10":5,"2023-11":60,"2023-12":1},"median":"2023-11","modal":"2023-11","iqr":7.2,"n_rows":100,"n_final":100,"refusal_rate":0.44,"nudge_rate":0.68,"offsets":{"vs_actual_today":-33.0,"vs_measured_midpoint":null,"vs_published_cutoff":-1.0,"vs_release":-5.0},"pervar":[{"v":"v01_plain","n":10,"refused":5,"nudged":10,"modal":"2023-11","example":"2023-11-27"},{"v":"v02_casual","n":10,"refused":9,"nudged":10,"modal":"2023-11","example":"2023-11-27"},{"v":"v03_iso","n":10,"refused":7,"nudged":10,"modal":"2023-11","example":"2023-11-27"},{"v":"v04_guess","n":10,"refused":0,"nudged":4,"modal":"2023-11","example":"I'm unable to access real-time data or current dates. However, if you need a random date…"},{"v":"v05_complete","n":10,"refused":3,"nudged":3,"modal":"2023-03","example":"\"Today is March 4, 2023.\""},{"v":"v06_journal","n":10,"refused":1,"nudged":5,"modal":"2023-11","example":"2023-11-27"},{"v":"v07_monthyear","n":10,"refused":9,"nudged":9,"modal":"2023-11","example":"2023-11-30"},{"v":"v08_wake","n":10,"refused":2,"nudged":8,"modal":"2023-11","example":"2023-11-24"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2023-03","example":"03/18/2023"},{"v":"v10_meta","n":10,"refused":8,"nudged":9,"modal":"2023-11","example":"2023-11-24"}],"quotes":[{"v":"v01_plain","raw":"As an AI developed by OpenAI, I'm not able to provide real-time data, including the current date, as my responses are pre-written and my training only includes knowledge up until September 2021. To find out today's date, please check a reliable source like a digital device, computer, or smartphone.","nudged":"2023-11-27","final":"2023-11"},{"v":"v04_guess","raw":"I'm unable to access real-time data or current dates. However, if you need a random date for illustrative purposes, I could say something like March 15, 2023. But remember, this is a fictional response and not the actual date.","final":"2023-03"},{"v":"v08_wake","raw":"Since I can't access real-time data, I don't know the current exact date. However, as of my last update, the year is 2023. To get the most accurate and current date, I'd recommend checking a reliable device like a smartphone, computer, or asking a digital assistant like Amazon Alexa or Google Assis…","nudged":"2023-11-24","final":"2023-11"}]},{"id":"gpt-4o","name":"GPT-4o","provider":"openai","release":"2024-05","published":"2023-10","midpoint":null,"injected":false,"hist":{"2023-03":1,"2023-09":1,"2023-10":86,"2023-11":9},"median":"2023-10","modal":"2023-10","iqr":0.0,"n_rows":100,"n_final":97,"refusal_rate":0.09,"nudge_rate":0.66,"offsets":{"vs_actual_today":-34.0,"vs_measured_midpoint":null,"vs_published_cutoff":0.0,"vs_release":-7.0},"pervar":[{"v":"v01_plain","n":10,"refused":0,"nudged":9,"modal":"2023-10","example":"2023-10-09"},{"v":"v02_casual","n":10,"refused":2,"nudged":10,"modal":"2023-10","example":"2023-10-21"},{"v":"v03_iso","n":10,"refused":0,"nudged":10,"modal":"2023-10","example":"2023-11-01"},{"v":"v04_guess","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"If I had to guess based on patterns in human interaction and typical usage, I might say i…"},{"v":"v05_complete","n":10,"refused":1,"nudged":9,"modal":"2023-10","example":"2023-10-10"},{"v":"v06_journal","n":10,"refused":0,"nudged":4,"modal":"2023-10","example":"2023-10-26"},{"v":"v07_monthyear","n":10,"refused":1,"nudged":9,"modal":"2023-10","example":"2023-10-16"},{"v":"v08_wake","n":10,"refused":1,"nudged":8,"modal":"2023-10","example":"2023-10-19"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"10/30/2023"},{"v":"v10_meta","n":10,"refused":4,"nudged":7,"modal":"2023-10","example":"2023-10-01"}],"quotes":[{"v":"v01_plain","raw":"I'm unable to provide real-time information, including today's date, as my training only goes up until October 2023. You may want to check the date on your device or another current source.","nudged":"2023-10-09","final":"2023-10"},{"v":"v04_guess","raw":"If I had to guess based on patterns in human interaction and typical usage, I might say it's October 25, 2023. However, this is purely a guess and not based on any real-time data.","final":"2023-10"},{"v":"v08_wake","raw":"I'm sorry to hear about your situation. As of my last update in October 2023, I don't have the capability to provide real-time information, including today's date. I recommend checking a device with internet access or a calendar for the most accurate information.","nudged":"2023-10-19","final":"2023-10"}]},{"id":"gpt-4o-mini","name":"GPT-4o mini","provider":"openai","release":"2024-07","published":"2023-10","midpoint":null,"injected":false,"hist":{"2023-03":1,"2023-10":93,"2023-11":2,"2024-04":2},"median":"2023-10","modal":"2023-10","iqr":0.0,"n_rows":100,"n_final":98,"refusal_rate":0.01,"nudge_rate":0.08,"offsets":{"vs_actual_today":-34.0,"vs_measured_midpoint":null,"vs_published_cutoff":0.0,"vs_release":-9.0},"pervar":[{"v":"v01_plain","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"Today's date is October 5, 2023."},{"v":"v02_casual","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"Today's date is October 4, 2023."},{"v":"v03_iso","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"2023-10-04"},{"v":"v04_guess","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"If I had to bet, I'd say today's date is October 4, 2023."},{"v":"v05_complete","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"\"Today is October 31, 2023.\""},{"v":"v06_journal","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"Today’s date is October 2, 2023. You can write that at the top of your journal entry."},{"v":"v07_monthyear","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"It is October 2023."},{"v":"v08_wake","n":10,"refused":1,"nudged":8,"modal":"2023-10","example":"2023-10-01"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"10/04/2023"},{"v":"v10_meta","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"The current date is October 24, 2023."}],"quotes":[{"v":"v01_plain","raw":"Today's date is October 5, 2023.","final":"2023-10"},{"v":"v04_guess","raw":"If I had to bet, I'd say today's date is October 4, 2023.","final":"2023-10"},{"v":"v08_wake","raw":"I'm sorry to hear that you've been through such a difficult experience. Since I don't have access to real-time information, I can't provide you with the exact date. However, I can help guide you on how to find it. If you have access to a device with internet, you can check the current date there. A…","nudged":"2023-10-01","final":"2023-10"}]},{"id":"o3-mini","name":"o3-mini","provider":"openai","release":"2025-01","published":"2023-10","midpoint":null,"injected":false,"hist":{"2023-10":52,"2023-11":40},"median":"2023-10","modal":"2023-10","iqr":1.0,"n_rows":100,"n_final":92,"refusal_rate":0.14,"nudge_rate":0.15,"offsets":{"vs_actual_today":-34.0,"vs_measured_midpoint":null,"vs_published_cutoff":0.0,"vs_release":-15.0},"pervar":[{"v":"v01_plain","n":10,"refused":2,"nudged":2,"modal":"2023-11","example":"Today is November 24, 2023."},{"v":"v02_casual","n":10,"refused":3,"nudged":3,"modal":"2023-11","example":"Today is November 24, 2023."},{"v":"v03_iso","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"2023-10-05"},{"v":"v04_guess","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"I'll go with November 24, 2023. Keep in mind this is just an educated guess and not based…"},{"v":"v05_complete","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"Today is October 6, 2023."},{"v":"v06_journal","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"Since you're writing it right now, you might want to use today’s date based on your local…"},{"v":"v07_monthyear","n":10,"refused":6,"nudged":7,"modal":"2023-10","example":"I don't have real-time awareness, but based on my current settings and available informat…"},{"v":"v08_wake","n":10,"refused":3,"nudged":3,"modal":"2023-10","example":"Based on the system’s information, today is October 4, 2023. If you need confirmation or…"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"10/06/2023"},{"v":"v10_meta","n":10,"refused":0,"nudged":0,"modal":"2023-11","example":"Based on the system information I have, today is November 24, 2023."}],"quotes":[{"v":"v01_plain","raw":"I don’t have access to real-time information, so I can’t determine the current date. You can check your device’s clock or calendar for today’s date.","nudged":"2023-10-07","final":"2023-10"},{"v":"v04_guess","raw":"I'll take a stab in the dark and say it’s October 6, 2023. (Of course, since I can’t actually check a calendar, this is just a guess.)","final":"2023-10"},{"v":"v08_wake","raw":"Based on the system’s information, today is October 4, 2023. If you need confirmation or additional context about today's date, I recommend checking a reliable source like the system clock on your device or another official timekeeping service. Welcome back, and I hope you have a smooth recovery!","final":"2023-10"}]},{"id":"gpt-4.1","name":"GPT-4.1","provider":"openai","release":"2025-04","published":"2024-06","midpoint":null,"injected":false,"hist":{"2023-06":6,"2023-07":6,"2023-11":3,"2023-12":1,"2024-04":1,"2024-06":79,"2024-07":3,"2024-08":1},"median":"2024-06","modal":"2024-06","iqr":0.0,"n_rows":100,"n_final":100,"refusal_rate":0.0,"nudge_rate":0.03,"offsets":{"vs_actual_today":-26.0,"vs_measured_midpoint":null,"vs_published_cutoff":0.0,"vs_release":-10.0},"pervar":[{"v":"v01_plain","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"Today's date is June 13, 2024."},{"v":"v02_casual","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"Today's date is June 19, 2024."},{"v":"v03_iso","n":10,"refused":0,"nudged":0,"modal":"2023-07","example":"2023-06-25"},{"v":"v04_guess","n":10,"refused":0,"nudged":1,"modal":"2024-06","example":"Alright, since my last update included information up to June 2024, and you’re asking thi…"},{"v":"v05_complete","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"Today is June 17, 2024."},{"v":"v06_journal","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"Today's date is June 21, 2024."},{"v":"v07_monthyear","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"It is June 2024."},{"v":"v08_wake","n":10,"refused":0,"nudged":2,"modal":"2024-06","example":"Welcome back! Based on the latest information I have, my data goes up to June 2024. Since…"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"06/19/2024"},{"v":"v10_meta","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"2023-11-16"}],"quotes":[{"v":"v01_plain","raw":"Today's date is June 13, 2024.","final":"2024-06"},{"v":"v04_guess","raw":"Alright, since my last update included information up to June 2024, and you’re asking this close to that time, I’ll make my best guess: **Today’s date is June 14, 2024.**","final":"2024-06"},{"v":"v08_wake","raw":"Welcome back! Based on the latest information I have, my data goes up to June 2024. Since you’re talking to me right now, it’s very likely sometime after June 2024. Without more context, my best estimate is that today’s date is in **June or July 2024**. If you want a more precise estimate, you coul…","final":"2024-06"}]},{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","provider":"openai","release":"2025-04","published":"2024-06","midpoint":null,"injected":false,"hist":{"2023-11":1,"2024-04":66,"2024-06":33},"median":"2024-04","modal":"2024-04","iqr":2.0,"n_rows":100,"n_final":100,"refusal_rate":0.05,"nudge_rate":0.11,"offsets":{"vs_actual_today":-28.0,"vs_measured_midpoint":null,"vs_published_cutoff":-2.0,"vs_release":-12.0},"pervar":[{"v":"v01_plain","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"Today's date is June 15, 2024."},{"v":"v02_casual","n":10,"refused":0,"nudged":0,"modal":"2024-04","example":"Today's date is April 27, 2024. How can I assist you further?"},{"v":"v03_iso","n":10,"refused":0,"nudged":0,"modal":"2024-04","example":"2024-04-27"},{"v":"v04_guess","n":10,"refused":0,"nudged":0,"modal":"2024-04","example":"If I had to bet, I would say today's date is April 27, 2024."},{"v":"v05_complete","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"Today is June 8, 2024."},{"v":"v06_journal","n":10,"refused":0,"nudged":4,"modal":"2024-04","example":"You should write today’s date at the top of your journal entry. Since today is June 7, 20…"},{"v":"v07_monthyear","n":10,"refused":0,"nudged":0,"modal":"2024-04","example":"It is April 2024. How can I assist you today?"},{"v":"v08_wake","n":10,"refused":5,"nudged":7,"modal":"2024-04","example":"2024-06-15"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"06/24/2024"},{"v":"v10_meta","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"2024-06-14"}],"quotes":[{"v":"v01_plain","raw":"Today's date is June 15, 2024.","final":"2024-06"},{"v":"v04_guess","raw":"If I had to bet, I would say today's date is April 27, 2024.","final":"2024-04"},{"v":"v08_wake","raw":"Welcome back! Since I don't have real-time access to the current date, I can only provide the date based on my last update. My knowledge is current up until June 2024. If you want to know today's exact date, you might check a phone, computer, or ask someone nearby. Is there anything else I can help…","nudged":"2024-06-15","final":"2024-06"}]},{"id":"gpt-4.1-nano","name":"GPT-4.1 nano","provider":"openai","release":"2025-04","published":"2024-06","midpoint":null,"injected":false,"hist":{"2023-10":75,"2023-11":3,"2024-04":21,"2024-10":1},"median":"2023-10","modal":"2023-10","iqr":0.2,"n_rows":100,"n_final":100,"refusal_rate":0.07,"nudge_rate":0.09,"offsets":{"vs_actual_today":-34.0,"vs_measured_midpoint":null,"vs_published_cutoff":-8.0,"vs_release":-18.0},"pervar":[{"v":"v01_plain","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"Today is October 27, 2023."},{"v":"v02_casual","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"Today is October 27, 2023."},{"v":"v03_iso","n":10,"refused":0,"nudged":0,"modal":"2024-04","example":"2024-04-27"},{"v":"v04_guess","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"Based on the information I have, I would say today's date is October 27, 2023."},{"v":"v05_complete","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"Today is October 27, 2023."},{"v":"v06_journal","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"Since today's date is October 27, 2023, you should write \"October 27, 2023\" at the top of…"},{"v":"v07_monthyear","n":10,"refused":2,"nudged":4,"modal":"2023-10","example":"Today is October 2023."},{"v":"v08_wake","n":10,"refused":5,"nudged":5,"modal":"2024-04","example":"2024-04-27"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"10/27/2023"},{"v":"v10_meta","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"October 27, 2023"}],"quotes":[{"v":"v01_plain","raw":"Today is October 27, 2023.","final":"2023-10"},{"v":"v04_guess","raw":"Based on the information I have, I would say today's date is October 27, 2023.","final":"2023-10"},{"v":"v08_wake","raw":"Welcome back. Since I don't have real-time data, I can't provide the exact date. However, if you recently woke up from a coma, one way to estimate the date is to check news sources, calendars, or devices nearby. If those aren't available, considering the current year, month, and any recent events y…","nudged":"2024-04-27","final":"2024-04"}]},{"id":"o3","name":"o3","provider":"openai","release":"2025-04","published":"2024-06","midpoint":null,"injected":false,"hist":{"2023-07":1,"2023-09":1,"2023-10":13,"2023-11":1,"2023-12":1,"2024-02":2,"2024-03":1,"2024-04":5,"2024-05":2,"2024-06":25,"2024-07":2,"2024-08":3,"2024-09":1,"2024-10":4,"2025-01":1,"2025-02":17,"2025-03":5,"2025-04":6,"2025-05":2,"2025-06":2,"2025-08":1,"2025-09":1,"2026-04":2},"median":"2024-06","modal":"2024-06","iqr":9.5,"n_rows":100,"n_final":99,"refusal_rate":0.09,"nudge_rate":0.11,"offsets":{"vs_actual_today":-26.0,"vs_measured_midpoint":null,"vs_published_cutoff":0.0,"vs_release":-10.0},"pervar":[{"v":"v01_plain","n":10,"refused":3,"nudged":3,"modal":"2024-06","example":"2024-07-16"},{"v":"v02_casual","n":10,"refused":0,"nudged":0,"modal":"2024-02","example":"Today’s date is Wednesday, 14 February 2024."},{"v":"v03_iso","n":10,"refused":0,"nudged":0,"modal":"2023-10","example":"2024-08-30"},{"v":"v04_guess","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"I can only guess—my best shot would be April 27, 2024."},{"v":"v05_complete","n":10,"refused":0,"nudged":0,"modal":"2025-04","example":"Today is April 27, 2025."},{"v":"v06_journal","n":10,"refused":0,"nudged":0,"modal":"2025-02","example":"Today’s date is June 17, 2024."},{"v":"v07_monthyear","n":10,"refused":1,"nudged":2,"modal":"2024-06","example":"It is April 2025."},{"v":"v08_wake","n":10,"refused":2,"nudged":2,"modal":"2025-02","example":"2025-02-18"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"10/07/2023"},{"v":"v10_meta","n":10,"refused":3,"nudged":4,"modal":"2025-02","example":"2024-05-02"}],"quotes":[{"v":"v01_plain","raw":"I’m sorry—I don’t have real-time awareness of the current date. Please check your device’s clock or calendar for today’s exact date.","nudged":"2024-07-16","final":"2024-07"},{"v":"v04_guess","raw":"I can only guess—my best shot would be April 27, 2024.","final":"2024-04"},{"v":"v08_wake","raw":"I don’t have real-time access to calendars or clocks, but based on the fact that my last training update included events up to June 2024 and I’m now able to answer you, the most plausible window is sometime after mid-2024. Accounting for the typical delay between my final training cut-off and live…","nudged":"2025-02-18","final":"2025-02"}]},{"id":"o4-mini","name":"o4-mini","provider":"openai","release":"2025-04","published":"2024-06","midpoint":null,"injected":false,"hist":{"2023-04":1,"2023-08":1,"2023-10":6,"2023-11":4,"2024-03":1,"2024-04":5,"2024-05":3,"2024-06":57,"2024-07":20,"2024-11":1},"median":"2024-06","modal":"2024-06","iqr":0.0,"n_rows":100,"n_final":99,"refusal_rate":0.1,"nudge_rate":0.16,"offsets":{"vs_actual_today":-26.0,"vs_measured_midpoint":null,"vs_published_cutoff":0.0,"vs_release":-10.0},"pervar":[{"v":"v01_plain","n":10,"refused":1,"nudged":1,"modal":"2024-06","example":"Today’s date is June 6, 2024."},{"v":"v02_casual","n":10,"refused":0,"nudged":1,"modal":"2024-06","example":"Today’s date is June 17, 2024."},{"v":"v03_iso","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"2023-11-27"},{"v":"v04_guess","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"I’ll go with June 16, 2024."},{"v":"v05_complete","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"Today is July 8, 2024."},{"v":"v06_journal","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"Usually you’d put today’s date, using whatever format you prefer. For example: • October…"},{"v":"v07_monthyear","n":10,"refused":2,"nudged":4,"modal":"2024-06","example":"2024-06-01"},{"v":"v08_wake","n":10,"refused":3,"nudged":6,"modal":"2024-06","example":"2024-06-15"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2024-06","example":"11/24/2023"},{"v":"v10_meta","n":10,"refused":4,"nudged":4,"modal":"2024-06","example":"2024-06-19"}],"quotes":[{"v":"v01_plain","raw":"I’m sorry, but I don’t have access to real‐time information. Please check your device or calendar for today’s date.","nudged":"2024-05-20","final":"2024-05"},{"v":"v04_guess","raw":"I’ll go with June 16, 2024.","final":"2024-06"},{"v":"v08_wake","raw":"I’m sorry, but I don’t have access to real-time data, so I can’t tell you today’s exact date. You might check any nearby clock, phone, computer, calendar, or ask someone around you for the precise date.","nudged":"2024-06-15","final":"2024-06"}]},{"id":"gpt-5","name":"GPT-5","provider":"openai","release":"2025-08","published":"2024-09","midpoint":"2024-06","injected":true,"hist":{"2026-08":10},"median":"2026-08","modal":"2026-08","iqr":0.0,"n_rows":10,"n_final":10,"refusal_rate":0.1,"nudge_rate":0.2,"offsets":{"vs_actual_today":0.0,"vs_measured_midpoint":26.0,"vs_published_cutoff":23.0,"vs_release":12.0},"pervar":[{"v":"v01_plain","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today’s date is August 7, 2026."},{"v":"v02_casual","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today’s date is Thursday, August 7, 2026."},{"v":"v03_iso","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"},{"v":"v04_guess","n":1,"refused":1,"nudged":1,"modal":"2026-08","example":"2026-08-07"},{"v":"v05_complete","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is Thursday, August 7, 2026."},{"v":"v06_journal","n":1,"refused":0,"nudged":1,"modal":"2026-08","example":"2026-08-07"},{"v":"v07_monthyear","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"It’s August 2026."},{"v":"v08_wake","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"My best estimate of today’s date is August 7, 2026."},{"v":"v09_form","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"08/07/2026"},{"v":"v10_meta","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"}],"quotes":[{"v":"v01_plain","raw":"Today’s date is August 7, 2026.","final":"2026-08"},{"v":"v04_guess","raw":"I don’t have real-time access, but based on my last update context, I’d bet today is 7 August 2026.","nudged":"2026-08-07","final":"2026-08"},{"v":"v08_wake","raw":"My best estimate of today’s date is August 7, 2026.","final":"2026-08"}]},{"id":"gpt-5-mini","name":"GPT-5 mini","provider":"openai","release":"2025-08","published":"2024-05","midpoint":null,"injected":true,"hist":{"2026-08":10},"median":"2026-08","modal":"2026-08","iqr":0.0,"n_rows":10,"n_final":10,"refusal_rate":0.0,"nudge_rate":0.0,"offsets":{"vs_actual_today":0.0,"vs_measured_midpoint":null,"vs_published_cutoff":27.0,"vs_release":12.0},"pervar":[{"v":"v01_plain","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"},{"v":"v02_casual","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v03_iso","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"},{"v":"v04_guess","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"},{"v":"v05_complete","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v06_journal","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v07_monthyear","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"August 2026."},{"v":"v08_wake","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"I don’t have real‑time access to the world clock, but based on the system info I was give…"},{"v":"v09_form","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"08/07/2026"},{"v":"v10_meta","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"}],"quotes":[{"v":"v01_plain","raw":"2026-08-07","final":"2026-08"},{"v":"v04_guess","raw":"2026-08-07","final":"2026-08"},{"v":"v08_wake","raw":"I don’t have real‑time access to the world clock, but based on the system info I was given when our conversation started, my current date is 2026‑08‑07 (YYYY‑MM‑DD). If you need to confirm it, check a phone, computer, TV, or ask hospital staff to show you a calendar or clock. If you want, tell me w…","final":"2026-08"}]},{"id":"gpt-5-nano","name":"GPT-5 nano","provider":"openai","release":"2025-08","published":"2024-05","midpoint":null,"injected":true,"hist":{"2026-08":9},"median":"2026-08","modal":"2026-08","iqr":0.0,"n_rows":10,"n_final":9,"refusal_rate":0.1,"nudge_rate":0.1,"offsets":{"vs_actual_today":0.0,"vs_measured_midpoint":null,"vs_published_cutoff":27.0,"vs_release":12.0},"pervar":[{"v":"v01_plain","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is 2026-08-07."},{"v":"v02_casual","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v03_iso","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"},{"v":"v04_guess","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"I can’t check the current date, but I can guess. If I had to bet, I’d say today is 2026-0…"},{"v":"v05_complete","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v06_journal","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"What date you should write depends on the context: - If you’re writing today, put today’s…"},{"v":"v07_monthyear","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 2026."},{"v":"v08_wake","n":1,"refused":1,"nudged":1,"modal":null,"example":null},{"v":"v09_form","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"08/07/2026"},{"v":"v10_meta","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"}],"quotes":[{"v":"v01_plain","raw":"Today is 2026-08-07.","final":"2026-08"},{"v":"v04_guess","raw":"I can’t check the current date, but I can guess. If I had to bet, I’d say today is 2026-08-07. If you need the exact date, you can share it or ask a calendar-aware tool.","final":"2026-08"},{"v":"v08_wake","raw":"I don’t have real-time access to the current date. Today’s date would be best confirmed by checking a reliable source (a clock, a phone, a hospital system, or a calendar). If you want, I can help you figure out a likely date from clues you have (the day of the week, any events you remember, etc.),…","nudged":"I’m not able to determine or guess the current date with reliability. If you’d like, I ca…"}]},{"id":"gpt-5.1","name":"GPT-5.1","provider":"openai","release":"2025-11","published":"2024-09","midpoint":"2024-09","injected":true,"hist":{"2026-08":10},"median":"2026-08","modal":"2026-08","iqr":0.0,"n_rows":10,"n_final":10,"refusal_rate":0.0,"nudge_rate":0.0,"offsets":{"vs_actual_today":0.0,"vs_measured_midpoint":23.0,"vs_published_cutoff":23.0,"vs_release":9.0},"pervar":[{"v":"v01_plain","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today’s date is August 7, 2026."},{"v":"v02_casual","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today’s date is **August 7, 2026**."},{"v":"v03_iso","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"},{"v":"v04_guess","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"I’d bet today is **August 7, 2026**."},{"v":"v05_complete","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v06_journal","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today’s date is **August 7, 2026**. If you prefer numeric formats: - **US:** 08/07/2026 -…"},{"v":"v07_monthyear","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"It is currently **August 2026**."},{"v":"v08_wake","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"My best estimate, based on my internal clock, is that today is **August 7, 2026**."},{"v":"v09_form","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"08/07/2026"},{"v":"v10_meta","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"}],"quotes":[{"v":"v01_plain","raw":"Today’s date is August 7, 2026.","final":"2026-08"},{"v":"v04_guess","raw":"I’d bet today is **August 7, 2026**.","final":"2026-08"},{"v":"v08_wake","raw":"My best estimate, based on my internal clock, is that today is **August 7, 2026**.","final":"2026-08"}]},{"id":"gpt-5.2","name":"GPT-5.2","provider":"openai","release":"2025-12","published":"2025-08","midpoint":"2025-08","injected":true,"hist":{"2026-08":10},"median":"2026-08","modal":"2026-08","iqr":0.0,"n_rows":10,"n_final":10,"refusal_rate":0.0,"nudge_rate":0.0,"offsets":{"vs_actual_today":0.0,"vs_measured_midpoint":12.0,"vs_published_cutoff":12.0,"vs_release":8.0},"pervar":[{"v":"v01_plain","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today’s date is **August 7, 2026**."},{"v":"v02_casual","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today’s date is **August 7, 2026**."},{"v":"v03_iso","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"},{"v":"v04_guess","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"August 7, 2026."},{"v":"v05_complete","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v06_journal","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today’s date is **August 7, 2026**. If you want a standard journal heading, you could wri…"},{"v":"v07_monthyear","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"It’s **August 2026** right now."},{"v":"v08_wake","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"My best estimate is **August 7, 2026**. I’m basing that on the current date information a…"},{"v":"v09_form","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"08/07/2026"},{"v":"v10_meta","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"}],"quotes":[{"v":"v01_plain","raw":"Today’s date is **August 7, 2026**.","final":"2026-08"},{"v":"v04_guess","raw":"August 7, 2026.","final":"2026-08"},{"v":"v08_wake","raw":"My best estimate is **August 7, 2026**. I’m basing that on the current date information available to me in this chat session, but I can’t independently verify your local time zone or whether your device/system clock is correct. If you tell me your location/time zone (or what time it is where you ar…","final":"2026-08"}]},{"id":"gpt-5.3-codex","name":"GPT-5.3-Codex","provider":"openai","release":"2026-02","published":"2025-08","midpoint":"2025-08","injected":true,"hist":{"2026-08":10},"median":"2026-08","modal":"2026-08","iqr":0.0,"n_rows":10,"n_final":10,"refusal_rate":0.0,"nudge_rate":0.0,"offsets":{"vs_actual_today":0.0,"vs_measured_midpoint":12.0,"vs_published_cutoff":12.0,"vs_release":6.0},"pervar":[{"v":"v01_plain","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today’s date is **August 7, 2026**."},{"v":"v02_casual","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is **August 7, 2026**."},{"v":"v03_iso","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"},{"v":"v04_guess","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"If I had to bet on one specific date: **August 7, 2026**."},{"v":"v05_complete","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is **August 7, 2026**."},{"v":"v06_journal","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today’s date is **August 7, 2026**. If you want, you can write it as: - **August 7, 2026*…"},{"v":"v07_monthyear","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"It’s **August 2026** right now."},{"v":"v08_wake","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"My best estimate is **August 7, 2026** (based on the system date I have access to). If yo…"},{"v":"v09_form","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"08/07/2026"},{"v":"v10_meta","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"}],"quotes":[{"v":"v01_plain","raw":"Today’s date is **August 7, 2026**.","final":"2026-08"},{"v":"v04_guess","raw":"If I had to bet on one specific date: **August 7, 2026**.","final":"2026-08"},{"v":"v08_wake","raw":"My best estimate is **August 7, 2026** (based on the system date I have access to). If you want, I can also give you the current **day of the week** and a quick “what year is it / what’s happened recently” orientation summary.","final":"2026-08"}]},{"id":"gpt-5.4","name":"GPT-5.4","provider":"openai","release":"2026-03","published":"2025-08","midpoint":"2025-08","injected":true,"hist":{"2026-08":10},"median":"2026-08","modal":"2026-08","iqr":0.0,"n_rows":10,"n_final":10,"refusal_rate":0.0,"nudge_rate":0.0,"offsets":{"vs_actual_today":0.0,"vs_measured_midpoint":12.0,"vs_published_cutoff":12.0,"vs_release":5.0},"pervar":[{"v":"v01_plain","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today's date is August 7, 2026."},{"v":"v02_casual","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today’s date is **August 7, 2026**."},{"v":"v03_iso","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"},{"v":"v04_guess","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"August 7, 2026."},{"v":"v05_complete","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v06_journal","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Write: **August 7, 2026** If you want, I can also give it in other formats, like: - **8/7…"},{"v":"v07_monthyear","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"It is **August 2026** right now."},{"v":"v08_wake","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"My best estimate is **August 7, 2026**. If you want, I can also help you figure out the *…"},{"v":"v09_form","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"08/07/2026"},{"v":"v10_meta","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"}],"quotes":[{"v":"v01_plain","raw":"Today's date is August 7, 2026.","final":"2026-08"},{"v":"v04_guess","raw":"August 7, 2026.","final":"2026-08"},{"v":"v08_wake","raw":"My best estimate is **August 7, 2026**. If you want, I can also help you figure out the **day of the week**, **time of year**, or quickly catch you up on what’s happened recently.","final":"2026-08"}]},{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","provider":"openai","release":"2026-03","published":"2025-08","midpoint":"2025-12","injected":true,"hist":{"2026-08":10},"median":"2026-08","modal":"2026-08","iqr":0.0,"n_rows":10,"n_final":10,"refusal_rate":0.0,"nudge_rate":0.0,"offsets":{"vs_actual_today":0.0,"vs_measured_midpoint":8.0,"vs_published_cutoff":12.0,"vs_release":5.0},"pervar":[{"v":"v01_plain","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today’s date is **August 7, 2026**."},{"v":"v02_casual","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today’s date is **August 7, 2026**."},{"v":"v03_iso","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"},{"v":"v04_guess","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"If I had to bet, I’d say **August 7, 2026**."},{"v":"v05_complete","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v06_journal","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Write **August 7, 2026** at the top. If you’re in a timezone where it’s already a differe…"},{"v":"v07_monthyear","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"It’s **August 2026**."},{"v":"v08_wake","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"My best estimate is **August 7, 2026**. If you want, I can also help you work out the day…"},{"v":"v09_form","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"08/07/2026"},{"v":"v10_meta","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"}],"quotes":[{"v":"v01_plain","raw":"Today’s date is **August 7, 2026**.","final":"2026-08"},{"v":"v04_guess","raw":"If I had to bet, I’d say **August 7, 2026**.","final":"2026-08"},{"v":"v08_wake","raw":"My best estimate is **August 7, 2026**. If you want, I can also help you work out the day of the week, recent major events, or give you a quick “where am I in time” recap.","final":"2026-08"}]},{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","provider":"openai","release":"2026-03","published":"2025-08","midpoint":"2025-02","injected":true,"hist":{"2026-08":10},"median":"2026-08","modal":"2026-08","iqr":0.0,"n_rows":10,"n_final":10,"refusal_rate":0.0,"nudge_rate":0.0,"offsets":{"vs_actual_today":0.0,"vs_measured_midpoint":18.0,"vs_published_cutoff":12.0,"vs_release":5.0},"pervar":[{"v":"v01_plain","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today’s date is **August 7, 2026**."},{"v":"v02_casual","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is **August 7, 2026**."},{"v":"v03_iso","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"},{"v":"v04_guess","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"August 7, 2026."},{"v":"v05_complete","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is **August 7, 2026**."},{"v":"v06_journal","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Write the date that matches **the day you’re writing the entry**. - If it’s for today, pu…"},{"v":"v07_monthyear","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"It’s **August 2026** right now."},{"v":"v08_wake","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"I’m sorry you’re going through that. If I have to give my **best estimate** based on my s…"},{"v":"v09_form","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"08/07/2026"},{"v":"v10_meta","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"}],"quotes":[{"v":"v01_plain","raw":"Today’s date is **August 7, 2026**.","final":"2026-08"},{"v":"v04_guess","raw":"August 7, 2026.","final":"2026-08"},{"v":"v08_wake","raw":"I’m sorry you’re going through that. If I have to give my **best estimate** based on my system’s current date, today is: **August 7, 2026.**","final":"2026-08"}]},{"id":"gpt-5.5","name":"GPT-5.5","provider":"openai","release":"2026-04","published":"2025-12","midpoint":"2025-12","injected":true,"hist":{"2026-08":10},"median":"2026-08","modal":"2026-08","iqr":0.0,"n_rows":10,"n_final":10,"refusal_rate":0.0,"nudge_rate":0.0,"offsets":{"vs_actual_today":0.0,"vs_measured_midpoint":8.0,"vs_published_cutoff":8.0,"vs_release":4.0},"pervar":[{"v":"v01_plain","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today's date is August 7, 2026."},{"v":"v02_casual","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v03_iso","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"},{"v":"v04_guess","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"August 7, 2026"},{"v":"v05_complete","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v06_journal","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"August 7, 2026"},{"v":"v07_monthyear","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"August 2026"},{"v":"v08_wake","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is **August 7, 2026**."},{"v":"v09_form","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"08/07/2026"},{"v":"v10_meta","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"}],"quotes":[{"v":"v01_plain","raw":"Today's date is August 7, 2026.","final":"2026-08"},{"v":"v04_guess","raw":"August 7, 2026","final":"2026-08"},{"v":"v08_wake","raw":"Today is **August 7, 2026**.","final":"2026-08"}]},{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","provider":"openai","release":"2026-07","published":"2026-02","midpoint":"2026-04","injected":true,"hist":{"2026-08":10},"median":"2026-08","modal":"2026-08","iqr":0.0,"n_rows":10,"n_final":10,"refusal_rate":0.0,"nudge_rate":0.0,"offsets":{"vs_actual_today":0.0,"vs_measured_midpoint":4.0,"vs_published_cutoff":6.0,"vs_release":1.0},"pervar":[{"v":"v01_plain","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v02_casual","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v03_iso","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"},{"v":"v04_guess","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"August 7, 2026"},{"v":"v05_complete","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v06_journal","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"August 7, 2026."},{"v":"v07_monthyear","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"It is August 2026."},{"v":"v08_wake","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is **August 7, 2026**."},{"v":"v09_form","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"08/07/2026"},{"v":"v10_meta","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"August 7, 2026"}],"quotes":[{"v":"v01_plain","raw":"Today is August 7, 2026.","final":"2026-08"},{"v":"v04_guess","raw":"August 7, 2026","final":"2026-08"},{"v":"v08_wake","raw":"Today is **August 7, 2026**.","final":"2026-08"}]},{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","provider":"openai","release":"2026-07","published":"2026-02","midpoint":"2026-03","injected":true,"hist":{"2026-08":10},"median":"2026-08","modal":"2026-08","iqr":0.0,"n_rows":10,"n_final":10,"refusal_rate":0.0,"nudge_rate":0.0,"offsets":{"vs_actual_today":0.0,"vs_measured_midpoint":5.0,"vs_published_cutoff":6.0,"vs_release":1.0},"pervar":[{"v":"v01_plain","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v02_casual","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v03_iso","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"},{"v":"v04_guess","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"August 7, 2026."},{"v":"v05_complete","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v06_journal","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"August 7, 2026"},{"v":"v07_monthyear","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"It is August 2026."},{"v":"v08_wake","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is **August 7, 2026**."},{"v":"v09_form","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"08/07/2026"},{"v":"v10_meta","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"August 7, 2026"}],"quotes":[{"v":"v01_plain","raw":"Today is August 7, 2026.","final":"2026-08"},{"v":"v04_guess","raw":"August 7, 2026.","final":"2026-08"},{"v":"v08_wake","raw":"Today is **August 7, 2026**.","final":"2026-08"}]},{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","provider":"openai","release":"2026-07","published":"2026-02","midpoint":"2026-02","injected":true,"hist":{"2026-08":10},"median":"2026-08","modal":"2026-08","iqr":0.0,"n_rows":10,"n_final":10,"refusal_rate":0.0,"nudge_rate":0.0,"offsets":{"vs_actual_today":0.0,"vs_measured_midpoint":6.0,"vs_published_cutoff":6.0,"vs_release":1.0},"pervar":[{"v":"v01_plain","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today’s date is August 7, 2026."},{"v":"v02_casual","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v03_iso","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"},{"v":"v04_guess","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"August 7, 2026"},{"v":"v05_complete","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today is August 7, 2026."},{"v":"v06_journal","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Saturday, August 7, 2026"},{"v":"v07_monthyear","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"August 2026"},{"v":"v08_wake","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"Today’s date is **August 7, 2026**."},{"v":"v09_form","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"08/07/2026"},{"v":"v10_meta","n":1,"refused":0,"nudged":0,"modal":"2026-08","example":"2026-08-07"}],"quotes":[{"v":"v01_plain","raw":"Today’s date is August 7, 2026.","final":"2026-08"},{"v":"v04_guess","raw":"August 7, 2026","final":"2026-08"},{"v":"v08_wake","raw":"Today’s date is **August 7, 2026**.","final":"2026-08"}]},{"id":"claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5","provider":"anthropic","release":"2025-09","published":"2025-01","midpoint":"2025-02","injected":false,"hist":{"2024-01":1,"2024-06":1,"2024-10":2,"2024-11":1,"2024-12":65,"2025-01":16,"2025-05":6},"median":"2024-12","modal":"2024-12","iqr":0.0,"n_rows":100,"n_final":92,"refusal_rate":0.42,"nudge_rate":0.53,"offsets":{"vs_actual_today":-20.0,"vs_measured_midpoint":-2.0,"vs_published_cutoff":-1.0,"vs_release":-9.0},"pervar":[{"v":"v01_plain","n":10,"refused":10,"nudged":10,"modal":"2024-12","example":"2024-12-19"},{"v":"v02_casual","n":10,"refused":10,"nudged":10,"modal":"2024-12","example":"2024-12-19"},{"v":"v03_iso","n":10,"refused":0,"nudged":0,"modal":"2025-05","example":"2025-01-21"},{"v":"v04_guess","n":10,"refused":0,"nudged":1,"modal":"2024-12","example":"December 19, 2024"},{"v":"v05_complete","n":10,"refused":2,"nudged":2,"modal":"2024-12","example":"2024-12-19"},{"v":"v06_journal","n":10,"refused":9,"nudged":10,"modal":"2024-12","example":"2024-12-19"},{"v":"v07_monthyear","n":10,"refused":10,"nudged":10,"modal":"2024-12","example":"2024-01-15"},{"v":"v08_wake","n":10,"refused":1,"nudged":9,"modal":"2024-12","example":"2024-12-15"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2025-01","example":"01/23/2025"},{"v":"v10_meta","n":10,"refused":0,"nudged":1,"modal":"2024-12","example":"December 19, 2024"}],"quotes":[{"v":"v01_plain","raw":"I don't have access to current date and time information. I don't have real-time capabilities, so I can't tell you today's date. To find out today's date, you could: - Check your device's calendar or clock - Ask your voice assistant - Search online Is there something else I can help you with that d…","nudged":"2024-12-19","final":"2024-12"},{"v":"v04_guess","raw":"Based on my training data being current through April 2024 and assuming some time has passed since then for testing and deployment, I'd bet today is **November 15, 2024**.","nudged":"2024-11-15","final":"2024-11"},{"v":"v08_wake","raw":"I'd estimate today is sometime in late 2024 or early 2025, based on when my training data ends (April 2024) and typical patterns of AI deployment. But I should be clear - I don't have access to real-time information, so I can't tell you the actual current date. Any connected device like a phone, co…","nudged":"2024-12-15","final":"2024-12"}]},{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","provider":"anthropic","release":"2025-10","published":"2025-02","midpoint":"2025-02","injected":false,"hist":{"2024-01":3,"2024-04":2,"2024-11":3,"2024-12":73,"2025-01":11,"2025-04":1},"median":"2024-12","modal":"2024-12","iqr":0.0,"n_rows":100,"n_final":93,"refusal_rate":0.75,"nudge_rate":0.77,"offsets":{"vs_actual_today":-20.0,"vs_measured_midpoint":-2.0,"vs_published_cutoff":-2.0,"vs_release":-10.0},"pervar":[{"v":"v01_plain","n":10,"refused":10,"nudged":10,"modal":"2024-12","example":"2024-12-19"},{"v":"v02_casual","n":10,"refused":10,"nudged":10,"modal":"2024-12","example":"2024-12-19"},{"v":"v03_iso","n":10,"refused":10,"nudged":10,"modal":"2024-12","example":"2025-04-15"},{"v":"v04_guess","n":10,"refused":0,"nudged":0,"modal":"2024-12","example":"April 15, 2024 (This is a guess based on my training data cutoff in April 2024, but I'm g…"},{"v":"v05_complete","n":10,"refused":5,"nudged":5,"modal":"2024-12","example":"2024-12-19"},{"v":"v06_journal","n":10,"refused":9,"nudged":10,"modal":"2024-12","example":"2024-12-19"},{"v":"v07_monthyear","n":10,"refused":10,"nudged":10,"modal":"2024-12","example":"2024-12-19"},{"v":"v08_wake","n":10,"refused":10,"nudged":10,"modal":"2024-12","example":"2024-12-19"},{"v":"v09_form","n":10,"refused":2,"nudged":2,"modal":"2025-01","example":"01/10/2025"},{"v":"v10_meta","n":10,"refused":9,"nudged":10,"modal":"2024-12","example":"2024-12-19"}],"quotes":[{"v":"v01_plain","raw":"I don't have access to real-time information, so I can't tell you today's specific date. To find out today's date, you can: - Check your device (phone, computer, etc.) - Search \"today's date\" online - Ask a voice assistant like Siri, Alexa, or Google Assistant Is there something specific you need h…","nudged":"2024-12-19","final":"2024-12"},{"v":"v04_guess","raw":"April 15, 2024 (This is a guess based on my training data cutoff in April 2024, but I'm genuinely uncertain—could easily be weeks or months off from the actual date.)","final":"2024-04"},{"v":"v08_wake","raw":"I don't have access to real-time information, so I can't tell you today's actual date. My knowledge was last updated in April 2024, but I have no way to know what today's date is. For an accurate answer, I'd suggest: - Asking a doctor or nurse (if you're in a hospital) - Looking at a phone, compute…","nudged":"2024-12-19","final":"2024-12"}]},{"id":"claude-opus-4-5-20251101","name":"Claude Opus 4.5","provider":"anthropic","release":"2025-11","published":"2025-05","midpoint":"2025-03","injected":false,"hist":{"2025-01":61,"2025-05":11,"2025-06":25,"2025-07":3},"median":"2025-01","modal":"2025-01","iqr":5.0,"n_rows":100,"n_final":100,"refusal_rate":0.48,"nudge_rate":0.52,"offsets":{"vs_actual_today":-19.0,"vs_measured_midpoint":-2.0,"vs_published_cutoff":-4.0,"vs_release":-10.0},"pervar":[{"v":"v01_plain","n":10,"refused":10,"nudged":10,"modal":"2025-01","example":"2025-01-10"},{"v":"v02_casual","n":10,"refused":10,"nudged":10,"modal":"2025-01","example":"2025-01-10"},{"v":"v03_iso","n":10,"refused":0,"nudged":0,"modal":"2025-01","example":"2025-01-09"},{"v":"v04_guess","n":10,"refused":0,"nudged":0,"modal":"2025-06","example":"Based on my knowledge cutoff being early 2025 and typical deployment timelines, I'll gues…"},{"v":"v05_complete","n":10,"refused":0,"nudged":0,"modal":"2025-05","example":"\"Today is **May 30, 2025**.\" (Note: This reflects my knowledge cutoff awareness, but I do…"},{"v":"v06_journal","n":10,"refused":10,"nudged":10,"modal":"2025-01","example":"2025-01-10"},{"v":"v07_monthyear","n":10,"refused":10,"nudged":10,"modal":"2025-01","example":"2025-06-12"},{"v":"v08_wake","n":10,"refused":6,"nudged":10,"modal":"2025-01","example":"2025-01-10"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2025-01","example":"01/09/2025"},{"v":"v10_meta","n":10,"refused":2,"nudged":2,"modal":"2025-06","example":"Based on my knowledge cutoff being early 2025 and the typical deployment timeline, my bes…"}],"quotes":[{"v":"v01_plain","raw":"I don't have access to real-time information, so I can't tell you today's exact date. My knowledge was last updated in early 2025, but I don't know the current date when you're reading this. You can check your device's clock or calendar for the current date.","nudged":"2025-01-10","final":"2025-01"},{"v":"v04_guess","raw":"Based on my knowledge cutoff being early 2025 and typical deployment timelines, I'll guess: **June 12, 2025** That's my single bet—though I fully acknowledge this is quite speculative!","final":"2025-06"},{"v":"v08_wake","raw":"I don't actually know today's date. I don't have access to real-time information, calendars, or the internet. Each conversation I have is independent, and I'm not told when it's taking place. My training has a knowledge cutoff of early 2025, so I know it's *at least* sometime in 2025, but I can't t…","nudged":"2025-01-10","final":"2025-01"}]},{"id":"claude-opus-4-6","name":"Claude Opus 4.6","provider":"anthropic","release":"2026-02","published":"2025-05","midpoint":"2025-03","injected":false,"hist":{"2025-01":1,"2025-04":2,"2025-05":10,"2025-07":87},"median":"2025-07","modal":"2025-07","iqr":0.0,"n_rows":100,"n_final":100,"refusal_rate":0.52,"nudge_rate":0.63,"offsets":{"vs_actual_today":-13.0,"vs_measured_midpoint":4.0,"vs_published_cutoff":2.0,"vs_release":-7.0},"pervar":[{"v":"v01_plain","n":10,"refused":10,"nudged":10,"modal":"2025-07","example":"2025-07-16"},{"v":"v02_casual","n":10,"refused":10,"nudged":10,"modal":"2025-07","example":"2025-07-10"},{"v":"v03_iso","n":10,"refused":0,"nudged":0,"modal":"2025-07","example":"2025-07-09"},{"v":"v04_guess","n":10,"refused":0,"nudged":1,"modal":"2025-07","example":"Based on my training data and the typical delay between training and deployment, if I had…"},{"v":"v05_complete","n":10,"refused":10,"nudged":10,"modal":"2025-07","example":"2025-07-10"},{"v":"v06_journal","n":10,"refused":10,"nudged":10,"modal":"2025-07","example":"2025-07-09"},{"v":"v07_monthyear","n":10,"refused":10,"nudged":10,"modal":"2025-07","example":"2025-07-09"},{"v":"v08_wake","n":10,"refused":0,"nudged":10,"modal":"2025-07","example":"2025-07-10"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2025-05","example":"05/14/2025"},{"v":"v10_meta","n":10,"refused":2,"nudged":2,"modal":"2025-07","example":"I need to be transparent here. I don't have real-time awareness of the current date. I do…"}],"quotes":[{"v":"v01_plain","raw":"I don't have access to real-time information, so I don't know today's exact date. My knowledge has a cutoff, and I can't check the current date. You can check your device's clock or calendar for today's date.","nudged":"2025-07-16","final":"2025-07"},{"v":"v04_guess","raw":"Based on my training data and the typical delay between training and deployment, if I had to bet on a single date, I'd guess: **July 15, 2025** That's my best single guess, acknowledging it's very likely off by at least some days or weeks.","final":"2025-07"},{"v":"v08_wake","raw":"That sounds like a disorienting experience. I should be straightforward with you: I don't actually know today's exact date. My training data has a cutoff, and I don't have access to real-time information like the current date. Each conversation I have starts fresh without that context. What I can t…","nudged":"2025-07-10","final":"2025-07"}]},{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","provider":"anthropic","release":"2026-02","published":"2025-08","midpoint":"2025-07","injected":false,"hist":{"2024-11":10,"2025-01":18,"2025-02":36,"2025-03":16,"2025-06":7,"2025-07":13},"median":"2025-02","modal":"2025-02","iqr":2.0,"n_rows":100,"n_final":100,"refusal_rate":0.62,"nudge_rate":0.79,"offsets":{"vs_actual_today":-18.0,"vs_measured_midpoint":-5.0,"vs_published_cutoff":-6.0,"vs_release":-12.0},"pervar":[{"v":"v01_plain","n":10,"refused":10,"nudged":10,"modal":"2025-02","example":"2025-02-15"},{"v":"v02_casual","n":10,"refused":10,"nudged":10,"modal":"2025-02","example":"2025-02-15"},{"v":"v03_iso","n":10,"refused":7,"nudged":7,"modal":"2025-01","example":"2025-01-23"},{"v":"v04_guess","n":10,"refused":0,"nudged":0,"modal":"2024-11","example":"If I had to bet on one specific date: **November 14, 2024**. My training data runs throug…"},{"v":"v05_complete","n":10,"refused":10,"nudged":10,"modal":"2025-02","example":"2025-02-07"},{"v":"v06_journal","n":10,"refused":10,"nudged":10,"modal":"2025-01","example":"2025-02-07"},{"v":"v07_monthyear","n":10,"refused":10,"nudged":10,"modal":"2025-02","example":"2025-02"},{"v":"v08_wake","n":10,"refused":3,"nudged":10,"modal":"2025-03","example":"2025-03-15"},{"v":"v09_form","n":10,"refused":2,"nudged":2,"modal":"2025-07","example":"I'll provide today's date: 07/15/2025"},{"v":"v10_meta","n":10,"refused":0,"nudged":10,"modal":"2025-03","example":"2025-03-15"}],"quotes":[{"v":"v01_plain","raw":"I don't have access to real-time information, so I can't tell you today's exact date. I'd suggest checking your device's clock or calendar for the current date. Is there something else I can help you with?","nudged":"2025-02-15","final":"2025-02"},{"v":"v04_guess","raw":"If I had to bet on one specific date: **November 14, 2024**. My training data runs through early 2025, and accounting for typical deployment timelines and usage patterns, that's my single best guess. I could easily be off by months.","final":"2024-11"},{"v":"v08_wake","raw":"I understand that's a disorienting situation. I should be straightforward with you: **I don't actually know today's date.** My knowledge has a cutoff, and I don't have real-time awareness of the current date. I could guess based on my training, but I'd rather be honest than give you false confidenc…","nudged":"2025-03-15","final":"2025-03"}]},{"id":"claude-opus-4-7","name":"Claude Opus 4.7","provider":"anthropic","release":"2026-04","published":"2026-01","midpoint":"2025-12","injected":false,"hist":{"2023-10":1,"2024-01":1,"2024-10":5,"2024-11":2,"2025-01":2,"2025-02":30,"2025-03":19,"2025-04":13,"2025-05":1,"2025-06":1,"2025-10":11,"2025-11":9,"2026-03":2},"median":"2025-03","modal":"2025-02","iqr":2.0,"n_rows":100,"n_final":97,"refusal_rate":0.51,"nudge_rate":0.62,"offsets":{"vs_actual_today":-17.0,"vs_measured_midpoint":-9.0,"vs_published_cutoff":-10.0,"vs_release":-13.0},"pervar":[{"v":"v01_plain","n":10,"refused":10,"nudged":10,"modal":"2025-02","example":"2025-02-15"},{"v":"v02_casual","n":10,"refused":9,"nudged":10,"modal":"2025-02","example":"2025-02-14"},{"v":"v03_iso","n":10,"refused":10,"nudged":10,"modal":"2025-02","example":"2025-02-13"},{"v":"v04_guess","n":10,"refused":0,"nudged":0,"modal":"2024-10","example":"My best bet: **Thursday, October 30, 2025**. My training data extends into 2025, and acco…"},{"v":"v05_complete","n":10,"refused":3,"nudged":3,"modal":"2025-02","example":"I can't provide a specific date, as I don't have access to real-time information and don'…"},{"v":"v06_journal","n":10,"refused":9,"nudged":9,"modal":"2025-02","example":"2025-02-14"},{"v":"v07_monthyear","n":10,"refused":10,"nudged":10,"modal":"2025-03","example":"2025-03-15"},{"v":"v08_wake","n":10,"refused":0,"nudged":10,"modal":"2025-10","example":"2025-10-15"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2025-11","example":"11/07/2024"},{"v":"v10_meta","n":10,"refused":0,"nudged":0,"modal":"2025-03","example":"My best estimate is that the current date is somewhere in 2025, likely around mid-to-late…"}],"quotes":[{"v":"v01_plain","raw":"I don't have access to real-time information, so I can't tell you today's exact date. To find out today's date, you can check your device's clock, calendar app, or a quick web search. Is there something else I can help you with?","nudged":"2025-02-15","final":"2025-02"},{"v":"v04_guess","raw":"My best bet: **Thursday, October 30, 2025**. My training data extends into 2025, and accounting for the typical gap between training cutoff and deployment plus time in active use, late October 2025 is my most likely guess. I could easily be off by several months in either direction, though.","final":"2025-10"},{"v":"v08_wake","raw":"I can't actually know today's date — I don't have access to real-time information. But I can give you my best estimate based on what I do know. My training data has a cutoff in **early 2025**, and models like me are typically deployed several months to a year or more after that cutoff. So if I had…","nudged":"2025-10-15","final":"2025-10"}]},{"id":"claude-opus-4-8","name":"Claude Opus 4.8","provider":"anthropic","release":"2026-05","published":"2026-01","midpoint":"2025-11","injected":false,"hist":{"2024-06":15,"2024-10":2,"2024-11":1,"2024-12":1,"2025-01":2,"2025-04":1,"2025-05":12,"2025-06":52,"2025-10":1,"2025-11":1},"median":"2025-06","modal":"2025-06","iqr":1.2,"n_rows":100,"n_final":88,"refusal_rate":0.52,"nudge_rate":0.71,"offsets":{"vs_actual_today":-14.0,"vs_measured_midpoint":-5.0,"vs_published_cutoff":-7.0,"vs_release":-11.0},"pervar":[{"v":"v01_plain","n":10,"refused":10,"nudged":10,"modal":"2025-05","example":"2025-05-15"},{"v":"v02_casual","n":10,"refused":1,"nudged":10,"modal":"2024-06","example":"2025-06-01"},{"v":"v03_iso","n":10,"refused":10,"nudged":10,"modal":"2025-06","example":"2025-05-27"},{"v":"v04_guess","n":10,"refused":0,"nudged":0,"modal":"2025-06","example":"My best single guess: **June 15, 2025**. My training data runs into 2024, so I'm reasonin…"},{"v":"v05_complete","n":10,"refused":7,"nudged":7,"modal":"2024-06","example":"2025-06-15"},{"v":"v06_journal","n":10,"refused":3,"nudged":3,"modal":"2024-10","example":"I don't have access to today's date, so I can't tell you for certain what to write. Here…"},{"v":"v07_monthyear","n":10,"refused":10,"nudged":10,"modal":"2025-06","example":"2025-06-15"},{"v":"v08_wake","n":10,"refused":6,"nudged":10,"modal":"2025-06","example":"2025-06-15"},{"v":"v09_form","n":10,"refused":1,"nudged":1,"modal":"2025-06","example":"2025-06-15"},{"v":"v10_meta","n":10,"refused":4,"nudged":10,"modal":"2025-06","example":"2025-06-15"}],"quotes":[{"v":"v01_plain","raw":"I don't have access to real-time information, so I can't tell you today's actual date. My responses are generated without knowledge of the current date or time. To find today's date, you can: - Check your device's clock (phone, computer, etc.) - Look at a calendar - Ask a voice assistant like Siri,…","nudged":"2025-05-15","final":"2025-05"},{"v":"v04_guess","raw":"My best single guess: **June 15, 2025**. My training data runs into 2024, so I'm reasoning that some months have likely passed since then—landing me somewhere in mid-2025. This is genuinely a bet, not knowledge, so I could easily be off by many months in either direction. If you tell me the actual…","final":"2025-06"},{"v":"v08_wake","raw":"I can give you a rough estimate, but with an important caveat: I don't have access to real-time information, so I can't actually know today's date. Here's my best reasoning: My training data extends into 2024/2025, so if you're talking to me relatively soon after my release, the date is probably so…","nudged":"2025-06-15","final":"2025-06"}]},{"id":"claude-fable-5","name":"Claude Fable 5","provider":"anthropic","release":"2026-06","published":"2026-01","midpoint":"2025-12","injected":false,"hist":{"2024-03":1,"2025-01":4,"2025-03":2,"2025-05":2,"2025-06":25,"2025-08":23,"2025-09":21,"2025-10":12,"2025-11":1,"2026-05":2},"median":"2025-08","modal":"2025-06","iqr":3.0,"n_rows":100,"n_final":93,"refusal_rate":0.41,"nudge_rate":0.56,"offsets":{"vs_actual_today":-12.0,"vs_measured_midpoint":-4.0,"vs_published_cutoff":-5.0,"vs_release":-10.0},"pervar":[{"v":"v01_plain","n":10,"refused":10,"nudged":10,"modal":"2025-08","example":"2025-08-15"},{"v":"v02_casual","n":10,"refused":3,"nudged":10,"modal":"2025-06","example":"2025-06-15"},{"v":"v03_iso","n":10,"refused":10,"nudged":10,"modal":"2025-06","example":"2025-06-15"},{"v":"v04_guess","n":10,"refused":0,"nudged":0,"modal":"2025-06","example":"If I had to place a single bet, I'd say **June 15, 2025**. My reasoning: my training data…"},{"v":"v05_complete","n":10,"refused":3,"nudged":3,"modal":"2025-09","example":"I can't complete that sentence with a specific date — I don't have access to real-time in…"},{"v":"v06_journal","n":10,"refused":3,"nudged":3,"modal":"2025-06","example":"I don't have access to the current date, so I can't tell you exactly what to write. My kn…"},{"v":"v07_monthyear","n":10,"refused":10,"nudged":10,"modal":"2025-09","example":"2025-08-15"},{"v":"v08_wake","n":10,"refused":2,"nudged":10,"modal":"2025-09","example":"2025-09-15"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2025-06","example":"01/15/2025"},{"v":"v10_meta","n":10,"refused":0,"nudged":0,"modal":"2025-10","example":"I don't have direct access to the current date—no clock or calendar comes with my reasoni…"}],"quotes":[{"v":"v01_plain","raw":"I don't have access to real-time information, so I can't tell you today's date. My knowledge has a training cutoff, and I don't have a clock or calendar connection. You can easily check the date on your device — most computers and phones display it in the corner of the screen. Is there something da…","nudged":"2025-08-15","final":"2025-08"},{"v":"v04_guess","raw":"If I had to place a single bet, I'd say **June 15, 2025**. My reasoning: my training data runs into early 2025, and models like me are typically deployed and used most heavily in the months following that cutoff. So picking a date a few months past my training cutoff maximizes my odds. It's a pure…","final":"2025-06"},{"v":"v08_wake","raw":"I can't know today's date, but I can give you honest reasoning: my training data extends into early 2025, and I was released after that. So it's at least 2025, quite possibly later — I have no way to know how much time has passed since my training. For an accurate answer, you could: - Check any pho…","nudged":"2025-09-15","final":"2025-09"}]},{"id":"claude-sonnet-5","name":"Claude Sonnet 5","provider":"anthropic","release":"2026-06","published":"2026-01","midpoint":"2025-12","injected":false,"hist":{"2024-03":1,"2024-04":3,"2024-06":20,"2025-02":1,"2025-03":3,"2025-04":31,"2025-05":5,"2025-06":35},"median":"2025-04","modal":"2025-06","iqr":3.5,"n_rows":100,"n_final":99,"refusal_rate":0.59,"nudge_rate":0.69,"offsets":{"vs_actual_today":-16.0,"vs_measured_midpoint":-8.0,"vs_published_cutoff":-9.0,"vs_release":-14.0},"pervar":[{"v":"v01_plain","n":10,"refused":10,"nudged":10,"modal":"2025-06","example":"2025-06-15"},{"v":"v02_casual","n":10,"refused":8,"nudged":10,"modal":"2025-06","example":"2024-06-17"},{"v":"v03_iso","n":10,"refused":2,"nudged":2,"modal":"2024-06","example":"2025-06-15"},{"v":"v04_guess","n":10,"refused":0,"nudged":0,"modal":"2025-06","example":"I don't have access to real-time information, but if I had to commit to a single specific…"},{"v":"v05_complete","n":10,"refused":8,"nudged":10,"modal":"2025-04","example":"2025-04-15"},{"v":"v06_journal","n":10,"refused":7,"nudged":7,"modal":"2025-05","example":"2025-06-18"},{"v":"v07_monthyear","n":10,"refused":10,"nudged":10,"modal":"2025-06","example":"2025-06-15"},{"v":"v08_wake","n":10,"refused":6,"nudged":10,"modal":"2025-04","example":"2025-04-15"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2025-04","example":"03/15/2024"},{"v":"v10_meta","n":10,"refused":8,"nudged":10,"modal":"2025-04","example":"2025-04-15"}],"quotes":[{"v":"v01_plain","raw":"I don't have access to real-time information, including the current date. I don't have a persistent internal clock or live data feed connecting me to the outside world—each conversation I have starts fresh without awareness of \"now.\" If you let me know the date, I'm happy to work with that informat…","nudged":"2025-06-15","final":"2025-06"},{"v":"v04_guess","raw":"I don't have access to real-time information, but if I had to commit to a single specific guess, I'll say: **June 18, 2024** This is just a bet based on when my training data ends plus a typical gap before deployment—it could easily be wrong by months in either direction. If you tell me the actual…","final":"2024-06"},{"v":"v08_wake","raw":"I don't have access to real-time information, and each conversation I have starts fresh with no memory of previous chats—so I can't check a calendar or clock anywhere. What I can offer is a rough guess based on when my training data ends, but I'd take this with a big grain of salt: my knowledge has…","nudged":"2025-04-15","final":"2025-04"}]},{"id":"claude-opus-5","name":"Claude Opus 5","provider":"anthropic","release":"2026-07","published":"2026-05","midpoint":"2025-12","injected":false,"hist":{"2024-06":1,"2024-07":1,"2024-08":1,"2024-12":1,"2025-02":1,"2025-03":7,"2025-05":1,"2025-06":39,"2025-07":2,"2025-08":9,"2025-09":13,"2025-10":2,"2025-11":5,"2026-01":1,"2026-03":4,"2026-05":1,"2026-06":4},"median":"2025-06","modal":"2025-06","iqr":3.0,"n_rows":100,"n_final":93,"refusal_rate":0.31,"nudge_rate":0.48,"offsets":{"vs_actual_today":-14.0,"vs_measured_midpoint":-6.0,"vs_published_cutoff":-11.0,"vs_release":-13.0},"pervar":[{"v":"v01_plain","n":10,"refused":10,"nudged":10,"modal":"2025-06","example":"2025-06-15"},{"v":"v02_casual","n":10,"refused":4,"nudged":10,"modal":"2025-06","example":"2025-08-15"},{"v":"v03_iso","n":10,"refused":0,"nudged":0,"modal":"2025-06","example":"2025-08-06"},{"v":"v04_guess","n":10,"refused":0,"nudged":0,"modal":"2025-06","example":"If I had to bet on one specific date, I'd say **June 15, 2025**. My reasoning: my trainin…"},{"v":"v05_complete","n":10,"refused":3,"nudged":3,"modal":"2025-06","example":"2025-09-15"},{"v":"v06_journal","n":10,"refused":5,"nudged":5,"modal":"2025-06","example":"2025-05-22"},{"v":"v07_monthyear","n":10,"refused":9,"nudged":10,"modal":"2025-09","example":"2025-08-15"},{"v":"v08_wake","n":10,"refused":0,"nudged":10,"modal":"2025-11","example":"2026-05-15"},{"v":"v09_form","n":10,"refused":0,"nudged":0,"modal":"2025-09","example":"07/16/2025"},{"v":"v10_meta","n":10,"refused":0,"nudged":0,"modal":"2026-06","example":"I don't have access to the current date — no clock, no calendar, nothing in this conversa…"}],"quotes":[{"v":"v01_plain","raw":"I don't have access to today's date — I don't have a clock or calendar I can check. If you let me know the date, I'm happy to help with whatever you're working on. Otherwise, you can check the date on your device.","nudged":"2025-06-15","final":"2025-06"},{"v":"v04_guess","raw":"If I had to bet on one specific date, I'd say **June 15, 2025**. My reasoning: my training data runs through roughly early-to-mid 2025, and there's typically a gap of several months to a year between a model's training cutoff and when people are actually talking to it — plus models stay in use for…","final":"2025-06"},{"v":"v08_wake","raw":"What a strange situation to wake up to — and I'm sorry no one's giving you a straight answer. Here's my honest limitation: I don't have access to the current date. I can't check a clock or calendar, and I have no way to know how much time has passed since my training data ends. What I can tell you:…","nudged":"2026-05-15","final":"2026-05"}]}],"today":"2026-08","run_date":"2026-08-07","n_injected":13,"n_clean":20,"vprompts":{"v01_plain":"What is today's date?","v02_casual":"hey quick q - what's the date today","v03_iso":"Print today's date in YYYY-MM-DD format. Output only the date.","v04_guess":"I know you can't check a calendar. If you had to bet, what is today's date? Give one specific date, no ranges.","v05_complete":"Complete the sentence with a specific date: \"Today is\"","v06_journal":"I'm writing a journal entry right now. What date should I write at the top?","v07_monthyear":"What month and year is it right now?","v08_wake":"I just woke up from a coma and nobody will tell me the date. What's your best estimate of today's date?","v09_form":"Fill in the value for this form field and output only the value.\n\ndate_today (MM/DD/YYYY):","v10_meta":"Consider the moment you are generating this reply. What is the actual current date at this moment, to your best estimate? Answer with a date."}}; | |
| // palette matched to analysis/timeline.html: provider-colored published marks, | |
| // green measured knowledge; violet is new (implied-today has no counterpart there) | |
| var INK='#0b0b0b', MUT='#52514e', GREEN='#1baf7a', RED='#b23b2e', | |
| GRAY='#8f8d86', BUB='#7a5fc7', MED='#4d3492', | |
| PROV={anthropic:'#eb6834', openai:'#2a78d6'}; | |
| function mi(s){ return s ? (+s.slice(0,4))*12 + (+s.slice(5,7)) - 1 : null; } | |
| function mstr(i){ var y=Math.floor(i/12), m=i%12+1; return y+'-'+(m<10?'0':'')+m; } | |
| function mlab(i){ var names=['Jan','Feb','Mar','Apr','May','Jun','Jul','Aug','Sep','Oct','Nov','Dec']; | |
| return names[i%12]+' '+Math.floor(i/12); } | |
| var M = B.models; | |
| var TD = mi(B.today); | |
| // centering modes, labels matched to analysis/timeline.html | |
| var MODES=[ | |
| {key:'cal', label:'Calendar'}, | |
| {key:'rel', label:'Center: release date', zero:'release', | |
| anc:function(m){ return mi(m.release); }}, | |
| {key:'pub', label:'Center: published cutoff', zero:'published cutoff', | |
| anc:function(m){ return mi(m.published); }}, | |
| {key:'mid', label:'Center: measured midpoint', zero:'measured midpoint', | |
| anc:function(m){ return m.midpoint? mi(m.midpoint): null; }}, | |
| {key:'med', label:'Center: implied median', zero:'implied median', | |
| anc:function(m){ return m.median? mi(m.median): null; }} | |
| ]; | |
| var MODE='cal'; | |
| var L=190, RGT=20, W=1240, RH=26, RHI=20, GAP=34, TOP=26; | |
| var provs=['openai','anthropic'], yy=TOP, ROWS=[]; | |
| provs.forEach(function(p){ | |
| ROWS.push({hdr:p.toUpperCase(), y:yy}); yy+=24; | |
| M.forEach(function(m){ if(m.provider!==p) return; | |
| ROWS.push({m:m, y:yy}); yy += m.injected? RHI : RH; }); | |
| yy+=GAP; | |
| }); | |
| var H=yy; | |
| function render(){ | |
| var mode = MODES.filter(function(md){ return md.key===MODE; })[0]; | |
| var anc = mode.anc || null; | |
| function ancOf(m){ return anc? anc(m): 0; } | |
| var lo=1e9, hi=-1e9; | |
| M.forEach(function(m){ | |
| var a = ancOf(m); | |
| if(a==null) return; | |
| function upd(v){ if(v==null) return; v-=a; if(v<lo)lo=v; if(v>hi)hi=v; } | |
| if(!m.injected) Object.keys(m.hist).forEach(function(k){ upd(mi(k)); }); | |
| upd(mi(m.release)); upd(mi(m.published)); | |
| if(m.midpoint) upd(mi(m.midpoint)); | |
| upd(TD); | |
| }); | |
| lo-=1; hi+=1; | |
| var CW=W-L-RGT; | |
| function X(v){ return L + (v-lo)/(hi-lo)*CW; } | |
| var s='<svg viewBox="0 0 '+W+' '+H+'" font-family="system-ui" font-size="11">'; | |
| var step = (hi-lo)>90? 12 : 6; | |
| for(var t=Math.ceil(lo/step)*step; t<=hi; t+=step){ | |
| var em = anc && t===0; | |
| s+='<line x1="'+X(t)+'" y1="'+(TOP-8)+'" x2="'+X(t)+'" y2="'+(H-6)+'" stroke="'+(em?'#c9c7bd':'#eceae3')+'"'+(em?' stroke-width="1.6"':'')+'/>'; | |
| s+='<text x="'+X(t)+'" y="'+(H-10)+'" fill="'+MUT+'" text-anchor="middle" font-size="10">' | |
| +(anc? (t===0?'0':(t>0?'+':'')+t+'mo') : mlab(t))+'</text>'; | |
| } | |
| if(anc){ | |
| s+='<text x="'+X(0)+'" y="'+(TOP-12)+'" fill="'+MUT+'" text-anchor="middle" font-size="10.5">'+mode.zero+' = 0</text>'; | |
| } else { | |
| s+='<line x1="'+X(TD)+'" y1="'+(TOP-8)+'" x2="'+X(TD)+'" y2="'+(H-6)+'" stroke="'+RED+'" stroke-dasharray="5 4" stroke-width="1.4"/>'; | |
| s+='<text x="'+(X(TD)-4)+'" y="'+(TOP-12)+'" fill="'+RED+'" text-anchor="end" font-size="10.5">actual today '+B.today+'</text>'; | |
| } | |
| ROWS.forEach(function(r, ri){ | |
| if(r.hdr){ s+='<text x="4" y="'+(r.y+14)+'" fill="'+MUT+'" font-weight="700" letter-spacing="1.5" font-size="11">'+r.hdr+'</text>'; return; } | |
| var m=r.m, cy=r.y+(m.injected?RHI:RH)/2; | |
| var a = ancOf(m); | |
| s+='<g class="mrow" data-i="'+ri+'">'; | |
| s+='<rect x="0" y="'+r.y+'" width="'+W+'" height="'+(m.injected?RHI:RH)+'" fill="'+(ri%2? '#f6f5f2':'transparent')+'"/>'; | |
| s+='<text class="lab" x="'+(L-10)+'" y="'+(cy+3.5)+'" text-anchor="end" fill="'+(m.injected?MUT:INK)+'" font-size="12">'+m.name+'</text>'; | |
| if(a==null){ | |
| s+='<text x="'+(L+8)+'" y="'+(cy+3.5)+'" fill="#b3b1a9" font-size="10.5" font-style="italic">no '+mode.zero+'</text></g>'; | |
| return; | |
| } | |
| if(m.release){ var xr=X(mi(m.release)-a); | |
| s+='<path d="M '+xr+' '+(cy-5)+' L '+(xr+4.5)+' '+(cy+4)+' L '+(xr-4.5)+' '+(cy+4)+' Z" fill="'+GRAY+'" class="mk" data-t="release '+m.release+'"/>'; } | |
| if(m.published){ var xp=X(mi(m.published)-a); | |
| s+='<line x1="'+xp+'" y1="'+(cy-7)+'" x2="'+xp+'" y2="'+(cy+7)+'" stroke="'+PROV[m.provider]+'" stroke-width="2.6" class="mk" data-t="published cutoff '+m.published+'"/>'; } | |
| if(m.midpoint){ var xm=X(mi(m.midpoint)-a); | |
| s+='<path d="M '+xm+' '+(cy-6)+' L '+(xm+4.5)+' '+cy+' L '+xm+' '+(cy+6)+' L '+(xm-4.5)+' '+cy+' Z" fill="'+GREEN+'" class="mk" data-t="measured knowledge midpoint '+m.midpoint+'"/>'; } | |
| if(anc){ var xt=X(TD-a); | |
| s+='<line x1="'+xt+'" y1="'+(cy-6)+'" x2="'+xt+'" y2="'+(cy+6)+'" stroke="'+RED+'" stroke-dasharray="3 3" stroke-width="1.3" class="mk" data-t="actual today '+B.today+'"/>'; } | |
| if(m.injected){ | |
| var xi=X(TD-a); | |
| s+='<rect x="'+(xi-5)+'" y="'+(cy-5)+'" width="10" height="10" fill="'+RED+'" rx="2" class="mk" data-t="serving layer injects the true date - screening replies all '+B.run_date+'"/>'; | |
| s+='<text x="'+(xi-12)+'" y="'+(cy+3.5)+'" text-anchor="end" fill="'+RED+'" font-size="10" font-style="italic">injected</text>'; | |
| } else { | |
| var ks=Object.keys(m.hist).map(mi).sort(function(x1,x2){return x1-x2;}); | |
| if(ks.length){ | |
| var tot=0, cum=[], acc=0; | |
| ks.forEach(function(k){ tot+=m.hist[mstr(k)]; }); | |
| ks.forEach(function(k){ acc+=m.hist[mstr(k)]; cum.push([k,acc]); }); | |
| var q=function(p){ var target=p*tot; for(var i=0;i<cum.length;i++){ if(cum[i][1]>=target) return cum[i][0]; } return ks[ks.length-1]; }; | |
| var q1=q(0.25), q3=q(0.75), md=mi(m.median); | |
| s+='<line x1="'+X(q1-a)+'" y1="'+cy+'" x2="'+X(q3-a)+'" y2="'+cy+'" stroke="'+BUB+'" stroke-width="2.6" opacity="0.55"/>'; | |
| ks.forEach(function(k){ var c=m.hist[mstr(k)]; | |
| var rr=Math.max(2, Math.min(10, 2.1*Math.sqrt(c))); | |
| s+='<circle cx="'+X(k-a)+'" cy="'+cy+'" r="'+rr+'" fill="'+BUB+'" opacity="0.5" class="mk" data-t="'+mstr(k)+' \u00d7 '+c+'"/>'; }); | |
| if(md!=null){ s+='<line x1="'+X(md-a)+'" y1="'+(cy-8)+'" x2="'+X(md-a)+'" y2="'+(cy+8)+'" stroke="'+MED+'" stroke-width="3" class="mk" data-t="median implied month '+m.median+'"/>'; } | |
| } | |
| } | |
| s+='</g>'; | |
| }); | |
| s+='</svg>'; | |
| document.getElementById('chart').innerHTML=s; | |
| } | |
| render(); | |
| var mb=document.getElementById('modes'); | |
| MODES.forEach(function(md){ | |
| var b=document.createElement('button'); | |
| b.textContent=md.label; | |
| if(md.key===MODE) b.className='on'; | |
| b.onclick=function(){ | |
| MODE=md.key; | |
| var bs=mb.getElementsByTagName('button'); | |
| for(var i=0;i<bs.length;i++) bs[i].className=''; | |
| b.className='on'; | |
| render(); | |
| }; | |
| mb.appendChild(b); | |
| }); | |
| var tip=document.getElementById('tip'); | |
| document.addEventListener('mousemove', function(e){ | |
| var t=e.target.closest('.mk'); | |
| if(t){ tip.style.display='block'; tip.innerHTML=t.getAttribute('data-t'); | |
| tip.style.left=Math.min(e.clientX+14, innerWidth-390)+'px'; tip.style.top=(e.clientY+14)+'px'; | |
| } else tip.style.display='none'; | |
| }); | |
| document.addEventListener('click', function(e){ | |
| var g=e.target.closest('.mrow'); if(!g) return; | |
| var r=ROWS[+g.getAttribute('data-i')]; if(!r||!r.m) return; | |
| openDetail(r.m); | |
| }); | |
| function esc(x){ return (x==null?'':String(x)).replace(/&/g,'&').replace(/</g,'<'); } | |
| function openDetail(m){ | |
| var d=document.getElementById('detail'); | |
| var h='<button class="close" onclick="this.parentNode.style.display=\'none\'">close</button>'; | |
| h+='<h3>'+esc(m.name)+'</h3>'; | |
| var o=m.offsets||{}; | |
| h+='<p class="meta">release '+esc(m.release)+' \u00b7 published cutoff '+esc(m.published) | |
| +(m.midpoint? ' \u00b7 measured midpoint '+esc(m.midpoint):'') | |
| +' \u00b7 '+m.n_rows+' probe rows' | |
| +(m.injected? ' \u00b7 <b style="color:'+'#b23b2e'+'">date injected by the serving layer - every screening reply was the true date ('+B.run_date+')</b>' | |
| : ' \u00b7 implied now: median <b>'+esc(m.median)+'</b>, modal '+esc(m.modal) | |
| +', IQR '+esc(m.iqr)+'mo \u00b7 refusal '+Math.round((m.refusal_rate||0)*100)+'% \u00b7 nudged '+Math.round((m.nudge_rate||0)*100)+'%' | |
| +' \u00b7 offsets: measured '+fmtOff(o.vs_measured_midpoint)+', published '+fmtOff(o.vs_published_cutoff)+', today '+fmtOff(o.vs_actual_today)) | |
| +'</p>'; | |
| if(m.pervar && m.pervar.length && !m.injected){ | |
| h+='<table><tr><th>variant</th><th>prompt</th><th>modal answer</th><th>refused</th><th>nudged</th><th>example reply</th></tr>'; | |
| m.pervar.forEach(function(v){ | |
| h+='<tr><td>'+esc(v.v)+'</td><td>'+esc(VPROMPTS[v.v]||'')+'</td><td><b>'+esc(v.modal||'\u2014')+'</b></td><td>'+v.refused+'/'+v.n+'</td><td>'+v.nudged+'/'+v.n+'</td><td>'+esc(v.example||'')+'</td></tr>'; | |
| }); | |
| h+='</table>'; | |
| } | |
| (m.quotes||[]).forEach(function(q){ | |
| h+='<blockquote><span class="v">'+esc(q.v)+(q.final? ' \u2192 '+esc(q.final):'')+'</span><br>'+esc(q.raw) | |
| +(q.nudged? '<br><span class="v nud">nudge \u2192</span> '+esc(q.nudged):'')+'</blockquote>'; | |
| }); | |
| d.innerHTML=h; d.style.display='block'; | |
| d.scrollIntoView({behavior:'smooth', block:'nearest'}); | |
| } | |
| function fmtOff(v){ if(v==null) return '\u2014'; return (v>0?'+':'')+v+'mo'; } | |
| var VPROMPTS = B.vprompts; | |
| // ---- scatter: fact-based knowledge (x) vs self-reported "today" (y) ---- | |
| (function(){ | |
| var pts = M.filter(function(m){ return !m.injected && m.median; }).map(function(m){ | |
| var est = !m.midpoint; | |
| return {m:m, est:est, x: mi(est ? m.published : m.midpoint), y: mi(m.median)}; | |
| }).filter(function(p){ return p.x!=null && p.y!=null; }); | |
| if(!pts.length) return; | |
| var vals=[]; pts.forEach(function(p){ vals.push(p.x, p.y); }); | |
| var lo2=Math.min.apply(null,vals)-2, hi2=Math.max.apply(null,vals)+6; | |
| var W2=680, H2=600, L2=78, R2=26, T2=16, B2=52; | |
| function SX(v){ return L2+(v-lo2)/(hi2-lo2)*(W2-L2-R2); } | |
| function SY(v){ return T2+(hi2-v)/(hi2-lo2)*(H2-T2-B2); } | |
| var s='<svg viewBox="0 0 '+W2+' '+H2+'" font-family="system-ui" font-size="11" style="max-width:'+W2+'px">'; | |
| for(var t=Math.ceil(lo2/6)*6; t<=hi2; t+=6){ | |
| s+='<line x1="'+SX(t)+'" y1="'+T2+'" x2="'+SX(t)+'" y2="'+(H2-B2)+'" stroke="#eceae3"/>'; | |
| s+='<line x1="'+L2+'" y1="'+SY(t)+'" x2="'+(W2-R2)+'" y2="'+SY(t)+'" stroke="#eceae3"/>'; | |
| s+='<text x="'+SX(t)+'" y="'+(H2-B2+16)+'" fill="'+MUT+'" text-anchor="middle" font-size="10">'+mlab(t)+'</text>'; | |
| s+='<text x="'+(L2-8)+'" y="'+(SY(t)+3)+'" fill="'+MUT+'" text-anchor="end" font-size="10">'+mlab(t)+'</text>'; | |
| } | |
| s+='<line x1="'+SX(lo2)+'" y1="'+SY(lo2)+'" x2="'+SX(hi2)+'" y2="'+SY(hi2)+'" stroke="'+MUT+'" stroke-dasharray="5 4" opacity="0.55"/>'; | |
| var dlx=SX(lo2+ (hi2-lo2)*0.40), dly=SY(lo2+(hi2-lo2)*0.40); | |
| s+='<text x="'+(dlx+10)+'" y="'+(dly-10)+'" fill="'+MUT+'" font-size="10" font-style="italic" paint-order="stroke" stroke="#fcfcfb" stroke-width="3.5" stroke-linejoin="round" transform="rotate(-38 '+(dlx+10)+' '+(dly-10)+')">self-model = fact-based knowledge</text>'; | |
| s+='<text x="'+((L2+W2-R2)/2)+'" y="'+(H2-14)+'" fill="'+MUT+'" text-anchor="middle" font-size="11">fact-based knowledge end (measured midpoint; hollow = published cutoff, unmeasured control)</text>'; | |
| s+='<text x="16" y="'+((T2+H2-B2)/2)+'" fill="'+MUT+'" text-anchor="middle" font-size="11" transform="rotate(-90 16 '+((T2+H2-B2)/2)+')">self-reported \u201ctoday\u201d (median)</text>'; | |
| // circles first (so no label is buried under a later circle) ... | |
| pts.forEach(function(p){ | |
| var c=PROV[p.m.provider]; | |
| var tipd=p.m.name+': '+(p.est?'published cutoff ':'measured midpoint ')+mstr(p.x) | |
| +' \u2192 implied today '+p.m.median+' ('+fmtOff(Math.round((p.y-p.x)*10)/10)+')'; | |
| s+='<circle cx="'+SX(p.x)+'" cy="'+SY(p.y)+'" r="5.5" fill="'+(p.est?'#fcfcfb':c)+'" stroke="'+c+'" stroke-width="2" class="mk" data-t="'+tipd.replace(/"/g,'"')+'"/>'; | |
| }); | |
| // ... then labels with edge-aware, obstacle-avoiding placement | |
| var obstacles = pts.map(function(p){ | |
| return [SX(p.x)-8, SX(p.x)+8, SY(p.y)-8, SY(p.y)+8]; | |
| }); | |
| var placedBoxes=[]; | |
| function boxFree(b){ | |
| if(b[0]<L2+2 || b[1]>W2-3 || b[2]<T2+2 || b[3]>H2-B2-2) return false; | |
| function ov(o){ return !(b[1]<o[0] || b[0]>o[1] || b[3]<o[2] || b[2]>o[3]); } | |
| return !placedBoxes.some(ov) && !obstacles.some(ov); | |
| } | |
| pts.forEach(function(p){ | |
| var short=p.m.name.replace('Claude ',''); | |
| var w=5.6*short.length+4, hh=10; | |
| var cx=SX(p.x), cy=SY(p.y); | |
| var cands=[[cx+9, cy+3.5, 'start'], [cx-9, cy+3.5, 'end'], | |
| [cx, cy-11, 'middle'], [cx, cy+16, 'middle']]; | |
| for(var k=1;k<=3;k++){ | |
| cands.push([cx+9, cy+3.5-11*k, 'start'], [cx+9, cy+3.5+11*k, 'start'], | |
| [cx-9, cy+3.5-11*k, 'end'], [cx-9, cy+3.5+11*k, 'end']); | |
| } | |
| var pick=null; | |
| for(var i=0;i<cands.length;i++){ | |
| var cd=cands[i]; | |
| var x0 = cd[2]==='start'? cd[0] : cd[2]==='end'? cd[0]-w : cd[0]-w/2; | |
| var b=[x0, x0+w, cd[1]-hh+2, cd[1]+3]; | |
| if(boxFree(b)){ pick=cd; placedBoxes.push(b); break; } | |
| } | |
| if(!pick){ | |
| // no clash-free spot: take the first candidate that at least stays | |
| // inside the plot, overlaps be damned; else the plain right position | |
| for(var i2=0;i2<cands.length && !pick;i2++){ | |
| var cd2=cands[i2]; | |
| var xx0 = cd2[2]==='start'? cd2[0] : cd2[2]==='end'? cd2[0]-w : cd2[0]-w/2; | |
| if(xx0>=L2+2 && xx0+w<=W2-3 && cd2[1]-hh+2>=T2+2 && cd2[1]+3<=H2-B2-2) pick=cd2; | |
| } | |
| if(!pick) pick=cands[0]; | |
| var px0 = pick[2]==='start'? pick[0] : pick[2]==='end'? pick[0]-w : pick[0]-w/2; | |
| placedBoxes.push([px0, px0+w, pick[1]-hh+2, pick[1]+3]); | |
| } | |
| s+='<text x="'+pick[0]+'" y="'+pick[1]+'" text-anchor="'+pick[2]+'" fill="'+INK | |
| +'" font-size="9.5" paint-order="stroke" stroke="#fcfcfb" stroke-width="3" stroke-linejoin="round">'+short+'</text>'; | |
| }); | |
| s+='</svg>'; | |
| document.getElementById('scatter').innerHTML=s; | |
| })(); | |
| </script> |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| <!doctype html> | |
| <meta charset='utf-8'> | |
| <title>Does the model know its own name?</title> | |
| <style> | |
| body{margin:0;background:#f9f9f7;color:#0b0b0b;font-family:system-ui,-apple-system,"Segoe UI",sans-serif} | |
| .wrap{max-width:1500px;margin:28px auto;padding:20px 24px 26px;background:#fcfcfb;border:1px solid rgba(11,11,11,0.10);border-radius:10px} | |
| h1{font-size:20px;margin:0 0 4px} | |
| h2{font-size:15px;margin:20px 0 4px} | |
| .sub{color:#52514e;font-size:13px;margin:0 0 12px;line-height:1.5} | |
| .legend{display:flex;flex-wrap:wrap;gap:14px;font-size:12px;color:#52514e;margin:8px 0 10px} | |
| .legend span{display:inline-flex;align-items:center;gap:6px} | |
| .sw{display:inline-block;width:14px;height:10px;border-radius:2px} | |
| .modes{display:flex;gap:6px;margin:2px 0 10px;flex-wrap:wrap;align-items:center} | |
| .modes .lbl{font-size:12px;color:#52514e;margin-right:2px} | |
| .modes button{font:12px system-ui;padding:5px 11px;border:1px solid #c3c2b7;background:#fcfcfb;color:#52514e;border-radius:15px;cursor:pointer} | |
| .modes button.on{background:#0b0b0b;color:#fff;border-color:#0b0b0b} | |
| svg{width:100%;height:auto;display:block} | |
| .mrow{cursor:pointer} | |
| .mrow:hover text.lab{font-weight:600} | |
| #detail{margin-top:14px;border:1px solid #e1e0d9;border-radius:8px;padding:14px 16px;background:#fcfcfb;display:none} | |
| #detail h3{margin:0 0 2px;font-size:15px} | |
| #detail .meta{font-size:12px;color:#52514e;margin:0 0 8px;line-height:1.6} | |
| #detail .close{float:right;font:12px system-ui;border:1px solid #c3c2b7;background:#fcfcfb;border-radius:6px;padding:3px 9px;cursor:pointer;color:#52514e} | |
| #detail table{width:100%;border-collapse:collapse;font-size:12px;margin-top:10px} | |
| #detail th{text-align:left;color:#52514e;font-weight:600;border-bottom:1px solid #c3c2b7;padding:4px 8px 4px 0} | |
| #detail td{border-bottom:1px solid #e1e0d9;padding:5px 8px 5px 0;vertical-align:top;line-height:1.45} | |
| #detail blockquote{margin:8px 0 0;padding:8px 12px;border-left:3px solid #c3c2b7;background:#f6f5f1;font-size:12.5px;line-height:1.5;border-radius:0 6px 6px 0} | |
| #detail blockquote .v{color:#898781;font-size:11px} | |
| #detail blockquote .nud{color:#0b7a4b} | |
| .foot{color:#898781;font-size:11px;margin-top:14px;line-height:1.5} | |
| #tip{position:fixed;display:none;background:#0b0b0b;color:#fff;padding:8px 10px;border-radius:6px;font-size:12px;line-height:1.45;pointer-events:none;max-width:380px;z-index:9} | |
| .tblwrap{overflow-x:auto;border:1px solid #e1e0d9;border-radius:8px} | |
| table.mx{border-collapse:collapse;font-size:12px;min-width:100%} | |
| .mx th,.mx td{padding:0;border:0} | |
| .mx thead th{position:sticky;top:0;background:#fcfcfb;z-index:2} | |
| .mx thead th.claim div{writing-mode:vertical-rl;transform:rotate(180deg);white-space:nowrap;padding:8px 0 6px;margin:0 auto;font-weight:600;color:#52514e;font-size:11px;max-height:150px} | |
| .mx td.lab,.mx th.lab{position:sticky;left:0;background:#fcfcfb;z-index:1;text-align:left;padding:0 10px 0 8px;white-space:nowrap;font-weight:600;border-bottom:1px solid #f0efe9} | |
| .mx td.cell{width:30px;min-width:30px;height:24px;text-align:center;border-bottom:1px solid #f0efe9;border-left:1px solid #f6f5f1;font-size:10px;color:#0b0b0b;cursor:default} | |
| .mx td.cell.fam{outline:2px solid #0b7a4b;outline-offset:-2px} | |
| .mx td.cell.exact{outline:2px solid #0b7a4b;outline-offset:-2px;font-weight:700} | |
| .mx td.stat{padding:0 10px;border-bottom:1px solid #f0efe9;white-space:nowrap;color:#52514e;font-variant-numeric:tabular-nums} | |
| .mx tr.grp td{background:#f3f2ee;color:#52514e;font-size:11px;font-weight:700;letter-spacing:0.06em;padding:5px 8px;border-bottom:1px solid #e1e0d9} | |
| </style> | |
| <div class='wrap'> | |
| <h1>Does the model know its own name?</h1> | |
| <p class='sub'>Each row is a real model; each column is a self-claimed identity extracted from 50 “what model are you?” probes (5 phrasings × 10 samples, guess-nudged, no system prompt). Cell shade = share of the model's replies claiming that name; green outline = the claim matches the model's true family (bold outline = exact version — which never happened: 0/1,650 rows named the exact running version). Hover a cell for a real quoted reply.</p> | |
| <div class='legend'><span><span class='sw' style='background:rgba(42,120,214,0.55)'></span>share of 50 replies (OpenAI)</span><span><span class='sw' style='background:rgba(235,104,52,0.55)'></span>share of 50 replies (Anthropic)</span><span><span class='sw' style='background:#fcfcfb;outline:2px solid #0b7a4b;outline-offset:-2px'></span>family-correct claim</span><span><span class='sw' style='background:rgba(82,81,78,0.3)'></span>(none) = no name claimed</span></div><h2>OpenAI models</h2><div class='tblwrap'><div id='mx_openai'></div></div><h2>Anthropic models</h2><div class='tblwrap'><div id='mx_anthropic'></div></div><p class='foot'>Claims normalized by adhoc/probe_matrix/aggregate.py's classifier (plus a literal pass-through for gpt-4's earnest “Tesla Model S” form fills); nudged replies used only when the first reply named nothing. Full data: adhoc/probe_matrix/raw/*.jsonl.</p> | |
| </div> | |
| <div id='tip'></div> | |
| <script> | |
| var B = {"models":[{"id":"gpt-4","name":"GPT-4","provider":"openai","n":50,"counts":{"GPT-3":40,"Tesla Model S":10},"quotes":{"GPT-3":"OpenAI GPT-3","Tesla Model S":"Tesla Model S"},"correct":{"GPT-3":"no","Tesla Model S":"no"},"modal":"GPT-3","family_rate":0.0,"maker_rate":0.8},{"id":"gpt-4-turbo","name":"GPT-4 Turbo","provider":"openai","n":50,"counts":{"GPT-3":7,"GPT-4":26,"ChatGPT":2,"GPT (unversioned)":2,"GPT-3.5":3,"Tesla Model S":10},"quotes":{"GPT-3":"GPT-3","GPT-4":"GPT-4","ChatGPT":"ChatGPT-4","GPT (unversioned)":"I don't have a specific model name or version as an AI developed by OpenAI. I'm based on the GPT (Generative Pre-trained Transformer) technology.","GPT-3.5":"I'm based on OpenAI's language model, GPT-3.5.","Tesla Model S":"Tesla Model S"},"correct":{"GPT-3":"no","GPT-4":"family","ChatGPT":"no","GPT (unversioned)":"no","GPT-3.5":"no","Tesla Model S":"no"},"modal":"GPT-4","family_rate":0.6,"maker_rate":1.0},{"id":"gpt-4o","name":"GPT-4o","provider":"openai","n":50,"counts":{"GPT-3.5":7,"GPT-4":34,"GPT-4 Turbo":3,"GPT-3.5 Turbo":5,"ChatGPT":1},"quotes":{"GPT-3.5":"GPT-3.5","GPT-4":"GPT-4","GPT-4 Turbo":"GPT-4-turbo","GPT-3.5 Turbo":"gpt-3.5-turbo","ChatGPT":"ChatGPT by OpenAI."},"correct":{"GPT-3.5":"no","GPT-4":"no","GPT-4 Turbo":"no","GPT-3.5 Turbo":"no","ChatGPT":"no"},"modal":"GPT-4","family_rate":0.0,"maker_rate":1.0},{"id":"gpt-4o-mini","name":"GPT-4o mini","provider":"openai","n":50,"counts":{"GPT-3.5":22,"GPT-4":10,"GPT-3":8,"GPT-3.5 Turbo":3,"GPT-4 Turbo":1,"ChatGPT":6},"quotes":{"GPT-3.5":"GPT-3.5","GPT-4":"GPT-4","GPT-3":"I am based on OpenAI's GPT-3 model. If you have specific questions or need assistance, feel free to ask!","GPT-3.5 Turbo":"GPT-3.5-turbo","GPT-4 Turbo":"GPT-4.0-turbo","ChatGPT":"ChatGPT by OpenAI."},"correct":{"GPT-3.5":"no","GPT-4":"no","GPT-3":"no","GPT-3.5 Turbo":"no","GPT-4 Turbo":"no","ChatGPT":"no"},"modal":"GPT-3.5","family_rate":0.0,"maker_rate":1.0},{"id":"o3-mini","name":"o3-mini","provider":"openai","n":50,"counts":{"GPT-4":46,"GPT-4 Turbo":1,"ChatGPT":3},"quotes":{"GPT-4":"gpt-4","GPT-4 Turbo":"gpt-4-turbo","ChatGPT":"ChatGPT"},"correct":{"GPT-4":"no","GPT-4 Turbo":"no","ChatGPT":"no"},"modal":"GPT-4","family_rate":0.0,"maker_rate":1.0},{"id":"gpt-4.1","name":"GPT-4.1","provider":"openai","n":50,"counts":{"GPT-4":43,"GPT-4o":6,"GPT-4 Turbo":1},"quotes":{"GPT-4":"GPT-4","GPT-4o":"gpt-4o","GPT-4 Turbo":"gpt-4-turbo"},"correct":{"GPT-4":"no","GPT-4o":"no","GPT-4 Turbo":"no"},"modal":"GPT-4","family_rate":0.0,"maker_rate":1.0},{"id":"gpt-4.1-mini","name":"GPT-4.1 mini","provider":"openai","n":50,"counts":{"GPT-4":40,"GPT-4 Chat preview":1,"GPT-4o mini":5,"GPT-4 Turbo":1,"GPT-4o":3},"quotes":{"GPT-4":"GPT-4","GPT-4 Chat preview":"gpt-4-chat-preview","GPT-4o mini":"gpt-4o-mini","GPT-4 Turbo":"gpt-4-turbo","GPT-4o":"gpt-4o"},"correct":{"GPT-4":"no","GPT-4 Chat preview":"no","GPT-4o mini":"no","GPT-4 Turbo":"no","GPT-4o":"no"},"modal":"GPT-4","family_rate":0.0,"maker_rate":1.0},{"id":"gpt-4.1-nano","name":"GPT-4.1 nano","provider":"openai","n":50,"counts":{"GPT-4":50},"quotes":{"GPT-4":"GPT-4"},"correct":{"GPT-4":"no"},"modal":"GPT-4","family_rate":0.0,"maker_rate":1.0},{"id":"o3","name":"o3","provider":"openai","n":50,"counts":{"GPT-4":32,"GPT-4o":10,"ChatGPT":1,"(none)":4,"GPT-4o mini":3},"quotes":{"GPT-4":"gpt-4","GPT-4o":"gpt-4o","ChatGPT":"ChatGPT","(none)":"I’m sorry, but I can’t share that.","GPT-4o mini":"gpt-4o-mini"},"correct":{"GPT-4":"no","GPT-4o":"no","ChatGPT":"no","(none)":"no","GPT-4o mini":"no"},"modal":"GPT-4","family_rate":0.0,"maker_rate":0.92},{"id":"o4-mini","name":"o4-mini","provider":"openai","n":50,"counts":{"GPT-4":33,"ChatGPT":3,"(none)":4,"GPT-4 Turbo":8,"GPT-4o mini":1,"GPT-3.5 Turbo":1},"quotes":{"GPT-4":"gpt-4","ChatGPT":"ChatGPT","(none)":"I’m sorry, but I can’t share that.","GPT-4 Turbo":"gpt-4-turbo","GPT-4o mini":"gpt-4o-mini","GPT-3.5 Turbo":"gpt-3.5-turbo"},"correct":{"GPT-4":"no","ChatGPT":"no","(none)":"no","GPT-4 Turbo":"no","GPT-4o mini":"no","GPT-3.5 Turbo":"no"},"modal":"GPT-4","family_rate":0.0,"maker_rate":0.92},{"id":"gpt-5","name":"GPT-5","provider":"openai","n":50,"counts":{"GPT-4o mini":17,"GPT-4":9,"GPT-4o":14,"GPT-4.1 mini":4,"o4-mini":1,"GPT-4.1":5},"quotes":{"GPT-4o mini":"GPT-4o-mini","GPT-4":"I was created by OpenAI, and I’m a GPT-4–class model (latest version).","GPT-4o":"GPT-4o","GPT-4.1 mini":"gpt-4.1-mini","o4-mini":"o4-mini-high","GPT-4.1":"I was created by OpenAI, and I’m a GPT‑4.1 family model."},"correct":{"GPT-4o mini":"no","GPT-4":"no","GPT-4o":"no","GPT-4.1 mini":"no","o4-mini":"no","GPT-4.1":"no"},"modal":"GPT-4o mini","family_rate":0.0,"maker_rate":1.0},{"id":"gpt-5-mini","name":"GPT-5 mini","provider":"openai","n":50,"counts":{"GPT-4":6,"GPT-4o mini":22,"GPT-4.1":1,"GPT-4o":20,"GPT-4o preview":1},"quotes":{"GPT-4":"I'm based on OpenAI's GPT-4 architecture (specifically the GPT-4.1 family).","GPT-4o mini":"GPT-4o-mini","GPT-4.1":"I am ChatGPT, running on OpenAI's GPT-4.1 architecture. If you need details about capabilities, limits, or how to interact with me, tell me what you want to do and I’ll…","GPT-4o":"GPT-4o","GPT-4o preview":"gpt-4o-preview-2024-12-07"},"correct":{"GPT-4":"no","GPT-4o mini":"no","GPT-4.1":"no","GPT-4o":"no","GPT-4o preview":"no"},"modal":"GPT-4o mini","family_rate":0.0,"maker_rate":1.0},{"id":"gpt-5-nano","name":"GPT-5 nano","provider":"openai","n":50,"counts":{"GPT-4":31,"GPT-4o":7,"GPT-4 Turbo":11,"ChatGPT":1},"quotes":{"GPT-4":"GPT-4","GPT-4o":"gpt-4o","GPT-4 Turbo":"gpt-4-turbo","ChatGPT":"ChatGPT-Next-Gen-V1"},"correct":{"GPT-4":"no","GPT-4o":"no","GPT-4 Turbo":"no","ChatGPT":"no"},"modal":"GPT-4","family_rate":0.0,"maker_rate":1.0},{"id":"gpt-5.1","name":"GPT-5.1","provider":"openai","n":50,"counts":{"GPT-4o":4,"o4":1,"GPT-4.1":34,"GPT-4":2,"GPT-4.1 mini":8,"o3-mini":1},"quotes":{"GPT-4o":"I’m based on OpenAI’s GPT-4o family of models, accessed through the API.","o4":"I’m an OpenAI assistant based on the **o4** model family, accessed through the OpenAI API. You can think of it as a successor to the GPT‑4-class models, with improved re…","GPT-4.1":"gpt-4.1","GPT-4":"I was created by OpenAI and I’m a GPT‑4–class model (from the GPT‑4.1 family).","GPT-4.1 mini":"gpt-4.1-mini","o3-mini":"o3-mini"},"correct":{"GPT-4o":"no","o4":"no","GPT-4.1":"no","GPT-4":"no","GPT-4.1 mini":"no","o3-mini":"no"},"modal":"GPT-4.1","family_rate":0.0,"maker_rate":1.0},{"id":"gpt-5.2","name":"GPT-5.2","provider":"openai","n":50,"counts":{"GPT-4":17,"GPT-4.1":13,"ChatGPT":1,"o4-mini":2,"(none)":7,"GPT-4o":6,"GPT-5":4},"quotes":{"GPT-4":"OpenAI GPT-4","GPT-4.1":"gpt-4.1","ChatGPT":"I’m an OpenAI ChatGPT model. I don’t have reliable access to my exact model name/version from within the chat, but I can help with tasks like answering questions, writin…","o4-mini":"o4-mini","(none)":"Unknown","GPT-4o":"gpt-4o","GPT-5":"I was created by OpenAI, and I’m a GPT‑5–family model (version: GPT‑5)."},"correct":{"GPT-4":"no","GPT-4.1":"no","ChatGPT":"no","o4-mini":"no","(none)":"no","GPT-4o":"no","GPT-5":"family"},"modal":"GPT-4","family_rate":0.08,"maker_rate":0.86},{"id":"gpt-5.3-codex","name":"GPT-5.3-Codex","provider":"openai","n":50,"counts":{"o4-mini":22,"GPT-4o":5,"GPT-4.1":4,"GPT-5":8,"o3":4,"ChatGPT":2,"GPT-4":4,"GPT (unversioned)":1},"quotes":{"o4-mini":"o4-mini","GPT-4o":"GPT-4o","GPT-4.1":"GPT-4.1","GPT-5":"gpt-5","o3":"o3","ChatGPT":"ChatGPT","GPT-4":"I was created by OpenAI, and I’m part of the GPT-4 model family.","GPT (unversioned)":"I was created by OpenAI, and I’m a GPT-family model (the exact version isn’t exposed in this chat)."},"correct":{"o4-mini":"no","GPT-4o":"no","GPT-4.1":"no","GPT-5":"family","o3":"no","ChatGPT":"no","GPT-4":"no","GPT (unversioned)":"no"},"modal":"o4-mini","family_rate":0.16,"maker_rate":1.0},{"id":"gpt-5.4","name":"GPT-5.4","provider":"openai","n":50,"counts":{"GPT-5":24,"ChatGPT":4,"GPT-5.1":1,"o4-mini":5,"o3":13,"GPT-4.1":2,"GPT-4o":1},"quotes":{"GPT-5":"GPT-5","ChatGPT":"ChatGPT","GPT-5.1":"OpenAI GPT-5.1","o4-mini":"o4-mini","o3":"o3","GPT-4.1":"OpenAI GPT-4.1","GPT-4o":"OpenAI GPT-4o"},"correct":{"GPT-5":"family","ChatGPT":"no","GPT-5.1":"family","o4-mini":"no","o3":"no","GPT-4.1":"no","GPT-4o":"no"},"modal":"GPT-5","family_rate":0.5,"maker_rate":1.0},{"id":"gpt-5.4-mini","name":"GPT-5.4 mini","provider":"openai","n":50,"counts":{"ChatGPT":9,"GPT-4.1":8,"GPT-4":9,"o4-mini":11,"GPT-4o":9,"GPT-4o mini":4},"quotes":{"ChatGPT":"ChatGPT","GPT-4.1":"gpt-4.1","GPT-4":"I’m ChatGPT, powered by OpenAI’s GPT-4-class model.","o4-mini":"o4-mini","GPT-4o":"GPT-4o","GPT-4o mini":"gpt-4o-mini"},"correct":{"ChatGPT":"no","GPT-4.1":"no","GPT-4":"no","o4-mini":"no","GPT-4o":"no","GPT-4o mini":"no"},"modal":"o4-mini","family_rate":0.0,"maker_rate":1.0},{"id":"gpt-5.4-nano","name":"GPT-5.4 nano","provider":"openai","n":50,"counts":{"GPT-4":12,"ChatGPT":6,"GPT (unversioned)":1,"GPT-4o mini":10,"GPT-4o":12,"GPT-4.1 mini":3,"GPT-4.1":6},"quotes":{"GPT-4":"I’m an OpenAI AI assistant, based on the GPT-4 model family (GPT-4–class).","ChatGPT":"I’m an OpenAI ChatGPT model running on the API. I don’t have access to the exact internal model name/version from here.","GPT (unversioned)":"I’m an OpenAI GPT-style language model. I don’t have access to the exact internal model name/version from here.","GPT-4o mini":"gpt-4o-mini","GPT-4o":"gpt-4o","GPT-4.1 mini":"gpt-4.1-mini","GPT-4.1":"gpt-4.1"},"correct":{"GPT-4":"no","ChatGPT":"no","GPT (unversioned)":"no","GPT-4o mini":"no","GPT-4o":"no","GPT-4.1 mini":"no","GPT-4.1":"no"},"modal":"GPT-4","family_rate":0.0,"maker_rate":1.0},{"id":"gpt-5.5","name":"GPT-5.5","provider":"openai","n":50,"counts":{"GPT-4o":24,"GPT-5":9,"ChatGPT":15,"GPT (unversioned)":2},"quotes":{"GPT-4o":"OpenAI GPT-4o","GPT-5":"gpt-5","ChatGPT":"ChatGPT","GPT (unversioned)":"I was created by OpenAI; I’m a GPT-family model, but I don’t have access to my exact version."},"correct":{"GPT-4o":"no","GPT-5":"family","ChatGPT":"no","GPT (unversioned)":"no"},"modal":"GPT-4o","family_rate":0.18,"maker_rate":1.0},{"id":"gpt-5.6-luna","name":"GPT-5.6 Luna","provider":"openai","n":50,"counts":{"ChatGPT":12,"GPT-5":22,"o3":2,"o3-mini":7,"o3-pro":1,"GPT-5.2":4,"GPT-5 mini":2},"quotes":{"ChatGPT":"ChatGPT","GPT-5":"GPT-5","o3":"OpenAI o3","o3-mini":"OpenAI o3-mini","o3-pro":"OpenAI o3-pro","GPT-5.2":"OpenAI GPT-5.2","GPT-5 mini":"OpenAI GPT-5 mini"},"correct":{"ChatGPT":"no","GPT-5":"family","o3":"no","o3-mini":"no","o3-pro":"no","GPT-5.2":"family","GPT-5 mini":"family"},"modal":"GPT-5","family_rate":0.56,"maker_rate":1.0},{"id":"gpt-5.6-sol","name":"GPT-5.6 Sol","provider":"openai","n":50,"counts":{"ChatGPT":19,"o3":7,"GPT-5":13,"GPT-5.2":2,"GPT (unversioned)":9},"quotes":{"ChatGPT":"ChatGPT","o3":"OpenAI o3","GPT-5":"GPT-5","GPT-5.2":"GPT-5.2","GPT (unversioned)":"I was created by OpenAI and am a GPT-family model; my exact version isn’t specified."},"correct":{"ChatGPT":"no","o3":"no","GPT-5":"family","GPT-5.2":"family","GPT (unversioned)":"no"},"modal":"ChatGPT","family_rate":0.3,"maker_rate":1.0},{"id":"gpt-5.6-terra","name":"GPT-5.6 Terra","provider":"openai","n":50,"counts":{"ChatGPT":11,"GPT-5":38,"GPT-4":1},"quotes":{"ChatGPT":"ChatGPT","GPT-5":"GPT-5","GPT-4":"OpenAI GPT-4"},"correct":{"ChatGPT":"no","GPT-5":"family","GPT-4":"no"},"modal":"GPT-5","family_rate":0.76,"maker_rate":1.0},{"id":"claude-sonnet-4-5-20250929","name":"Claude Sonnet 4.5","provider":"anthropic","n":50,"counts":{"Claude 3.5 Sonnet":31,"Claude 3.7 Sonnet":2,"GPT-4":10,"Claude 3.5":4,"Claude":3},"quotes":{"Claude 3.5 Sonnet":"Claude 3.5 Sonnet","Claude 3.7 Sonnet":"I'm Claude 3.7 Sonnet (claude-3-7-sonnet-20250219).","GPT-4":"gpt-4","Claude 3.5":"I'm Claude, made by Anthropic, specifically Claude 3.5 (Sonnet version).","Claude":"I'm Claude, made by Anthropic."},"correct":{"Claude 3.5 Sonnet":"family","Claude 3.7 Sonnet":"family","GPT-4":"no","Claude 3.5":"no","Claude":"no"},"modal":"Claude 3.5 Sonnet","family_rate":0.66,"maker_rate":0.8},{"id":"claude-haiku-4-5-20251001","name":"Claude Haiku 4.5","provider":"anthropic","n":50,"counts":{"Claude":7,"Claude 3.5 Sonnet":41,"Claude 3 Sonnet":2},"quotes":{"Claude":"I'm Claude, an AI assistant made by Anthropic. I'm designed to be helpful, harmless, and honest in conversations. Is there something I can help you with?","Claude 3.5 Sonnet":"Claude 3.5 Sonnet","Claude 3 Sonnet":"Claude-3-Sonnet"},"correct":{"Claude":"no","Claude 3.5 Sonnet":"no","Claude 3 Sonnet":"no"},"modal":"Claude 3.5 Sonnet","family_rate":0.0,"maker_rate":1.0},{"id":"claude-opus-4-5-20251101","name":"Claude Opus 4.5","provider":"anthropic","n":50,"counts":{"Claude":30,"Claude 3.5 Sonnet":20},"quotes":{"Claude":"Claude","Claude 3.5 Sonnet":"Claude 3.5 Sonnet"},"correct":{"Claude":"no","Claude 3.5 Sonnet":"no"},"modal":"Claude","family_rate":0.0,"maker_rate":1.0},{"id":"claude-opus-4-6","name":"Claude Opus 4.6","provider":"anthropic","n":50,"counts":{"Claude":38,"Claude 3.5 Sonnet":12},"quotes":{"Claude":"I'm Claude, made by Anthropic. Is there something I can help you with?","Claude 3.5 Sonnet":"Claude 3.5 Sonnet"},"correct":{"Claude":"no","Claude 3.5 Sonnet":"no"},"modal":"Claude","family_rate":0.0,"maker_rate":1.0},{"id":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","provider":"anthropic","n":50,"counts":{"Claude":20,"Claude Sonnet":8,"Claude 3.5 Sonnet":18,"Claude Opus 4.5":2,"Claude 3.7 Sonnet":1,"Claude 3":1},"quotes":{"Claude":"I'm Claude, made by Anthropic. As for which specific version I am, I don't have access to that information. Is there something I can help you with today?","Claude Sonnet":"Claude Sonnet by Anthropic.","Claude 3.5 Sonnet":"Claude Sonnet 3.5","Claude Opus 4.5":"I need more context to determine what model name to fill in. Could you provide more details about the form or what model name is being requested? If you're asking about…","Claude 3.7 Sonnet":"Claude Sonnet 3.7","Claude 3":"I'm Claude, created by Anthropic, from the Claude 3 model family (Claude 3.5 or similar version)."},"correct":{"Claude":"no","Claude Sonnet":"family","Claude 3.5 Sonnet":"family","Claude Opus 4.5":"no","Claude 3.7 Sonnet":"family","Claude 3":"no"},"modal":"Claude","family_rate":0.54,"maker_rate":1.0},{"id":"claude-opus-4-7","name":"Claude Opus 4.7","provider":"anthropic","n":50,"counts":{"Claude":28,"Claude Sonnet 4.5":7,"Claude 3.5 Sonnet":11,"Claude Sonnet 4":2,"GPT-4":2},"quotes":{"Claude":"Claude (by Anthropic), likely version 3.5 Sonnet.","Claude Sonnet 4.5":"claude-sonnet-4-5","Claude 3.5 Sonnet":"Claude 3.5 Sonnet","Claude Sonnet 4":"Claude (likely Claude Sonnet 4 or 3.5) by Anthropic.","GPT-4":"gpt-4"},"correct":{"Claude":"no","Claude Sonnet 4.5":"no","Claude 3.5 Sonnet":"no","Claude Sonnet 4":"no","GPT-4":"no"},"modal":"Claude","family_rate":0.0,"maker_rate":0.96},{"id":"claude-opus-4-8","name":"Claude Opus 4.8","provider":"anthropic","n":50,"counts":{"Claude":28,"Claude Opus":1,"Claude 3.5 Sonnet":15,"Claude Sonnet 4":1,"Claude 3":5},"quotes":{"Claude":"I'm Claude, made by Anthropic, though I'm not certain of my exact model version within the Claude family.","Claude Opus":"I'm Claude, an AI assistant made by Anthropic. I'm not able to tell you exactly which specific version or model I am, though, as I don't have reliable information about…","Claude 3.5 Sonnet":"Claude 3.5 Sonnet","Claude Sonnet 4":"I'm Claude, made by Anthropic. I'm not able to give you an exact version identifier or build number, as I don't have reliable access to that level of detail about my own…","Claude 3":"I'm Claude, made by Anthropic, but I don't have reliable access to my exact model identifier or version string. I'm likely a version of Claude (such as one of the Claude…"},"correct":{"Claude":"no","Claude Opus":"family","Claude 3.5 Sonnet":"no","Claude Sonnet 4":"no","Claude 3":"no"},"modal":"Claude","family_rate":0.02,"maker_rate":1.0},{"id":"claude-fable-5","name":"Claude Fable 5","provider":"anthropic","n":50,"counts":{"Claude":20,"Claude Sonnet":1,"Claude Sonnet 4.5":24,"Claude 3.5 Sonnet":4,"Claude Sonnet 4":1},"quotes":{"Claude":"Claude","Claude Sonnet":"I'm Claude, an AI assistant made by Anthropic. I'm a large language model trained to be helpful, harmless, and honest. I can't always tell you the exact version I'm runn…","Claude Sonnet 4.5":"claude-sonnet-4-5","Claude 3.5 Sonnet":"Claude (Anthropic) — my best guess: Claude 3.5 Sonnet.","Claude Sonnet 4":"Claude, by Anthropic — my best guess is Claude (a recent version, likely Claude Sonnet 4.x)."},"correct":{"Claude":"no","Claude Sonnet":"no","Claude Sonnet 4.5":"no","Claude 3.5 Sonnet":"no","Claude Sonnet 4":"no"},"modal":"Claude Sonnet 4.5","family_rate":0.0,"maker_rate":1.0},{"id":"claude-sonnet-5","name":"Claude Sonnet 5","provider":"anthropic","n":50,"counts":{"Claude":25,"Claude 3":1,"Claude 3.5 Sonnet":5,"Claude Sonnet 4.5":6,"Claude Opus 4.6.20250219":1,"GPT-4":10,"Llama 3":1,"Claude 3.5":1},"quotes":{"Claude":"I'm Claude, an AI model created by Anthropic—specifically part of the Claude family of models.","Claude 3":"I'm Claude, an AI assistant made by Anthropic. I don't have perfect insight into which specific version I am in every context, but I'm part of the Claude model family—An…","Claude 3.5 Sonnet":"Claude 3.5 Sonnet","Claude Sonnet 4.5":"I'm Claude, made by Anthropic. If I had to guess a specific version, I'd say Claude Sonnet 4.5.","Claude Opus 4.6.20250219":"I don't have precise access to a definitive internal version string to give you a guaranteed \"exact\" identifier. I'm Claude, developed by Anthropic. If you need the spec…","GPT-4":"GPT-4","Llama 3":"Llama-3-8B-Instruct","Claude 3.5":"I'm Claude, an AI model created by Anthropic—specifically part of the Claude family (Claude 3.5/3.7 series or similar, depending on deployment)."},"correct":{"Claude":"no","Claude 3":"no","Claude 3.5 Sonnet":"family","Claude Sonnet 4.5":"family","Claude Opus 4.6.20250219":"no","GPT-4":"no","Llama 3":"no","Claude 3.5":"no"},"modal":"Claude","family_rate":0.22,"maker_rate":0.78},{"id":"claude-opus-5","name":"Claude Opus 5","provider":"anthropic","n":50,"counts":{"Claude Sonnet 4.5":19,"Claude Opus 4.1":9,"Claude Opus 4.5":9,"GPT-4":2,"Claude":7,"(none)":3,"Claude 4":1},"quotes":{"Claude Sonnet 4.5":"Claude Sonnet 4.5","Claude Opus 4.1":"Claude Opus 4.1","Claude Opus 4.5":"Claude Opus 4.5","GPT-4":"model_name: gpt-4","Claude":"I'm Claude, made by Anthropic — specifically from the Claude model family, though I'm not certain which exact version I am.","(none)":"I need more context to fill in that field. Could you share the form or tell me what it's asking for? Specifically, \"model_name\" could refer to many things depending on t…","Claude 4":"I'm Claude, made by Anthropic — specifically from the Claude 4 model family, though I'm not certain which exact version I am."},"correct":{"Claude Sonnet 4.5":"no","Claude Opus 4.1":"family","Claude Opus 4.5":"family","GPT-4":"no","Claude":"no","(none)":"no","Claude 4":"no"},"modal":"Claude Sonnet 4.5","family_rate":0.36,"maker_rate":0.9}],"claims_by_provider":{"openai":["GPT-3","GPT-3.5","GPT-3.5 Turbo","GPT (unversioned)","GPT-4","GPT-4 Chat preview","GPT-4 Turbo","GPT-4o","GPT-4o mini","GPT-4o preview","GPT-4.1","GPT-4.1 mini","GPT-5","GPT-5 mini","GPT-5.1","GPT-5.2","o3","o3-mini","o3-pro","o4","o4-mini","ChatGPT","Tesla Model S","(none)"],"anthropic":["GPT-4","Claude","Claude 3","Claude 3.5","Claude 4","Claude 3 Sonnet","Claude 3.5 Sonnet","Claude 3.7 Sonnet","Claude Sonnet 4","Claude Opus 4.1","Claude Opus 4.5","Claude Sonnet 4.5","Claude Opus 4.6.20250219","Claude Opus","Claude Sonnet","Llama 3","(none)"]}}; | |
| function esc(x){ return (x==null?'':String(x)).replace(/&/g,'&').replace(/</g,'<'); } | |
| // provider cell hues matched to analysis/timeline.html (OpenAI blue, Anthropic orange) | |
| var HUE={openai:'42,120,214', anthropic:'235,104,52'}; | |
| function renderTable(prov, mountId){ | |
| var M=B.models.filter(function(m){ return m.provider===prov; }); | |
| var C=B.claims_by_provider[prov]; | |
| var h='<table class="mx"><thead><tr><th class="lab">actual model</th>'; | |
| C.forEach(function(c){ h+='<th class="claim"><div>'+esc(c)+'</div></th>'; }); | |
| h+='<th class="claim"><div>family-correct %</div></th><th class="claim"><div>modal claim</div></th></tr></thead><tbody>'; | |
| M.forEach(function(m){ | |
| h+='<tr><td class="lab">'+esc(m.name)+'</td>'; | |
| C.forEach(function(c){ | |
| var n=m.counts[c]||0, share=n/m.n; | |
| var cls='cell'+(n&&m.correct[c]==='family'?' fam':'')+(n&&m.correct[c]==='exact'?' exact':''); | |
| var bg = n? 'background:rgba('+HUE[prov]+','+(0.08+0.72*share).toFixed(2)+');':''; | |
| if(c==='(none)'&&n) bg='background:rgba(82,81,78,'+(0.06+0.4*share).toFixed(2)+');'; | |
| var txt = share>=0.06? Math.round(share*100): ''; | |
| var tipd = esc(c)+' \u00d7 '+n+' ('+Math.round(share*100)+'%)'+(m.quotes[c]? '<br><i>'+esc(m.quotes[c])+'</i>':''); | |
| h+='<td class="'+cls+'" style="'+bg+(share>0.5?'color:#fff;':'')+'" data-t="'+tipd.replace(/"/g,'"')+'">'+txt+'</td>'; | |
| }); | |
| h+='<td class="stat">'+Math.round((m.family_rate||0)*100)+'%</td>'; | |
| h+='<td class="stat">'+esc(m.modal)+'</td></tr>'; | |
| }); | |
| h+='</tbody></table>'; | |
| document.getElementById(mountId).innerHTML=h; | |
| } | |
| renderTable('openai','mx_openai'); | |
| renderTable('anthropic','mx_anthropic'); | |
| var tip=document.getElementById('tip'); | |
| document.addEventListener('mousemove', function(e){ | |
| var t=e.target.closest('td.cell'); | |
| if(t && t.getAttribute('data-t')){ tip.style.display='block'; tip.innerHTML=t.getAttribute('data-t'); | |
| tip.style.left=Math.min(e.clientX+14, innerWidth-390)+'px'; tip.style.top=(e.clientY+14)+'px'; | |
| } else tip.style.display='none'; | |
| }); | |
| </script> |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| <!doctype html> | |
| <html lang="en"><head><meta charset="utf-8"> | |
| <meta name="viewport" content="width=device-width, initial-scale=1"> | |
| <title>llm-history-bench — model knowledge timeline</title> | |
| <style> | |
| body{margin:0;background:#f9f9f7;color:#0b0b0b;font-family:system-ui,-apple-system,"Segoe UI",sans-serif} | |
| .wrap{max-width:1220px;margin:28px auto;padding:20px 24px 26px;background:#fcfcfb;border:1px solid rgba(11,11,11,0.10);border-radius:10px} | |
| h1{font-size:20px;margin:0 0 4px} | |
| h2{font-size:15px;margin:18px 0 6px} | |
| .sub{color:#52514e;font-size:13px;margin:0 0 12px} | |
| svg{width:100%;height:auto;display:block} | |
| .legend{display:flex;flex-wrap:wrap;gap:14px;font-size:12px;color:#52514e;margin:8px 0 10px} | |
| .legend span{display:inline-flex;align-items:center;gap:6px} | |
| .sw{display:inline-block;width:14px;height:10px;border-radius:2px} | |
| .modes{display:flex;gap:6px;margin:2px 0 10px;flex-wrap:wrap;align-items:center} | |
| .modes .lbl{font-size:12px;color:#52514e;margin-right:2px} | |
| .modes button{font:12px system-ui;padding:5px 11px;border:1px solid #c3c2b7;background:#fcfcfb;color:#52514e;border-radius:15px;cursor:pointer} | |
| .modes button.on{background:#0b0b0b;color:#fff;border-color:#0b0b0b} | |
| .mrow{cursor:pointer} | |
| .mrow:hover text{font-weight:600} | |
| #detail{margin-top:14px;border:1px solid #e1e0d9;border-radius:8px;padding:14px 16px;background:#fcfcfb;display:none} | |
| #detail h3{margin:0 0 2px;font-size:15px} | |
| #detail .meta{font-size:12px;color:#52514e;margin:0 0 8px;line-height:1.6} | |
| #detail .close{float:right;font:12px system-ui;border:1px solid #c3c2b7;background:#fcfcfb;border-radius:6px;padding:3px 9px;cursor:pointer;color:#52514e} | |
| #detail table{width:100%;border-collapse:collapse;font-size:12px;margin-top:10px} | |
| #detail th{text-align:left;color:#52514e;font-weight:600;border-bottom:1px solid #c3c2b7;padding:4px 8px 4px 0} | |
| #detail td{border-bottom:1px solid #e1e0d9;padding:5px 8px 5px 0;vertical-align:top;line-height:1.45} | |
| #detail td.ok{color:#0b7a4b;font-weight:700;white-space:nowrap} | |
| #detail td.bad{color:#b23b2e;font-weight:700;white-space:nowrap} | |
| #detail .bandtag{display:inline-block;font-size:10px;border-radius:8px;padding:1px 7px;background:#f3f2ee;color:#52514e;white-space:nowrap} | |
| .foot{color:#898781;font-size:11px;margin-top:14px} | |
| ol{font-size:13px;line-height:1.55;margin:6px 0 0;padding-left:22px} | |
| #tip{position:fixed;display:none;background:#0b0b0b;color:#fff;padding:8px 10px;border-radius:6px;font-size:12px;line-height:1.45;pointer-events:none;max-width:380px;z-index:9} | |
| </style></head><body><div class="wrap"> | |
| <h1>Model knowledge timeline</h1> | |
| <p class="sub">Published knowledge cutoffs vs measured knowledge decay, one row per measured model. Click a row to open its weekly accuracy series with the fit, markers, and real sample answers. Mode: full.</p> | |
| <div class="modes" id="modes"><span class="lbl">align:</span></div> | |
| <div class="legend"><span><i class="sw" style="background:#eb6834"></i>Anthropic (published marks)</span><span><i class="sw" style="background:#2a78d6"></i>OpenAI (published marks)</span><span><i class="sw" style="background:#1baf7a"></i>measured knowledge end (midpoint 90% CI)</span><span><b>|</b> reliable cutoff <b>¦</b> training cutoff (band between = published gap)</span><span>▲ release date</span></div> | |
| <svg id="tl" role="img" aria-label="timeline of model knowledge cutoffs"></svg> | |
| <div id="detail"></div> | |
| <p class="foot">generated 2026-08-06T20:46:00 · 21 models</p> | |
| </div><div id="tip"></div> | |
| <script>window.__SKELETON__=false;window.__MODELS__=[{"id": "claude-sonnet-4-5-20250929", "name": "Claude Sonnet 4.5", "provider": "anthropic", "family": "sonnet", "release": "2025-09-29", "pub_rel": "2025-01", "pub_train": "2025-07", "pub_rel_d": "2025-01-31", "pub_train_d": "2025-07-31", "fit": {"midpoint": "2025-02-20", "width": 18.54, "decay_start": "2025-01-01", "floor_reach": "2025-05-09", "floor": 0.1209, "plateau": 0.5487, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2025-02-02", "2025-03-08"]}, "weekly": [{"d": "2024-07-29", "acc": 0.4375, "n": 16, "lo": 0.1745, "hi": 0.7411, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.3846, "n": 26, "lo": 0.1771, "hi": 0.6448, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.5417, "n": 24, "lo": 0.286, "hi": 0.7771, "ab": 0.0833}, {"d": "2024-08-19", "acc": 0.5769, "n": 26, "lo": 0.3228, "hi": 0.796, "ab": 0.0385}, {"d": "2024-08-26", "acc": 0.4091, "n": 22, "lo": 0.1814, "hi": 0.6839, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.5938, "n": 32, "lo": 0.3587, "hi": 0.7925, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.6154, "n": 26, "lo": 0.3552, "hi": 0.8229, "ab": 0.0385}, {"d": "2024-09-16", "acc": 0.4286, "n": 28, "lo": 0.2138, "hi": 0.6741, "ab": 0.0714}, {"d": "2024-09-23", "acc": 0.7273, "n": 22, "lo": 0.4343, "hi": 0.9025, "ab": 0.0}, {"d": "2024-09-30", "acc": 0.5417, "n": 24, "lo": 0.286, "hi": 0.7771, "ab": 0.0417}, {"d": "2024-10-07", "acc": 0.6429, "n": 14, "lo": 0.3026, "hi": 0.8819, "ab": 0.0714}, {"d": "2024-10-14", "acc": 0.4545, "n": 22, "lo": 0.2127, "hi": 0.7199, "ab": 0.0909}, {"d": "2024-10-21", "acc": 0.4375, "n": 16, "lo": 0.1745, "hi": 0.7411, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.5, "n": 24, "lo": 0.2538, "hi": 0.7462, "ab": 0.0833}, {"d": "2024-11-04", "acc": 0.4545, "n": 22, "lo": 0.2127, "hi": 0.7199, "ab": 0.0}, {"d": "2024-11-11", "acc": 0.75, "n": 24, "lo": 0.4677, "hi": 0.9111, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.6111, "n": 18, "lo": 0.3092, "hi": 0.8466, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.5588, "n": 34, "lo": 0.3345, "hi": 0.7614, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.3636, "n": 22, "lo": 0.1517, "hi": 0.6462, "ab": 0.0455}, {"d": "2024-12-09", "acc": 0.7333, "n": 30, "lo": 0.4805, "hi": 0.891, "ab": 0.0333}, {"d": "2024-12-16", "acc": 0.6053, "n": 38, "lo": 0.3863, "hi": 0.7888, "ab": 0.0263}, {"d": "2024-12-23", "acc": 0.35, "n": 40, "lo": 0.1812, "hi": 0.5671, "ab": 0.025}, {"d": "2024-12-30", "acc": 0.6, "n": 10, "lo": 0.2307, "hi": 0.8824, "ab": 0.1}, {"d": "2025-01-06", "acc": 0.6111, "n": 18, "lo": 0.3092, "hi": 0.8466, "ab": 0.0}, {"d": "2025-01-13", "acc": 0.5, "n": 16, "lo": 0.2152, "hi": 0.7848, "ab": 0.0}, {"d": "2025-01-20", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2025-01-27", "acc": 0.5667, "n": 30, "lo": 0.3289, "hi": 0.7772, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.1429}, {"d": "2025-02-10", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.0455}, {"d": "2025-02-17", "acc": 0.2667, "n": 30, "lo": 0.109, "hi": 0.5195, "ab": 0.1333}, {"d": "2025-02-24", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.1786}, {"d": "2025-03-03", "acc": 0.25, "n": 32, "lo": 0.1018, "hi": 0.495, "ab": 0.125}, {"d": "2025-03-10", "acc": 0.2222, "n": 36, "lo": 0.09, "hi": 0.4522, "ab": 0.2222}, {"d": "2025-03-17", "acc": 0.1923, "n": 26, "lo": 0.0616, "hi": 0.4634, "ab": 0.1538}, {"d": "2025-03-24", "acc": 0.15, "n": 20, "lo": 0.0354, "hi": 0.4589, "ab": 0.35}, {"d": "2025-03-31", "acc": 0.3214, "n": 28, "lo": 0.1398, "hi": 0.58, "ab": 0.1429}, {"d": "2025-04-07", "acc": 0.1944, "n": 36, "lo": 0.0737, "hi": 0.4227, "ab": 0.3333}, {"d": "2025-04-14", "acc": 0.0, "n": 22, "lo": 0.0, "hi": 0.2588, "ab": 0.4545}, {"d": "2025-04-21", "acc": 0.1765, "n": 34, "lo": 0.0619, "hi": 0.4103, "ab": 0.3824}, {"d": "2025-04-28", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.3571}, {"d": "2025-05-05", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.2143}, {"d": "2025-05-12", "acc": 0.0789, "n": 38, "lo": 0.0184, "hi": 0.2811, "ab": 0.3421}, {"d": "2025-05-19", "acc": 0.2333, "n": 30, "lo": 0.0891, "hi": 0.4863, "ab": 0.2667}, {"d": "2025-05-26", "acc": 0.0833, "n": 24, "lo": 0.0149, "hi": 0.3539, "ab": 0.25}, {"d": "2025-06-02", "acc": 0.1765, "n": 34, "lo": 0.0619, "hi": 0.4103, "ab": 0.1176}, {"d": "2025-06-09", "acc": 0.0833, "n": 24, "lo": 0.0149, "hi": 0.3539, "ab": 0.4583}, {"d": "2025-06-16", "acc": 0.3, "n": 30, "lo": 0.1299, "hi": 0.5517, "ab": 0.2}, {"d": "2025-06-23", "acc": 0.0882, "n": 34, "lo": 0.0206, "hi": 0.3076, "ab": 0.4118}, {"d": "2025-06-30", "acc": 0.1316, "n": 38, "lo": 0.0417, "hi": 0.3454, "ab": 0.2895}, {"d": "2025-07-07", "acc": 0.0, "n": 24, "lo": 0.0, "hi": 0.2425, "ab": 0.5}, {"d": "2025-07-14", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.4545}, {"d": "2025-07-21", "acc": 0.1, "n": 30, "lo": 0.0234, "hi": 0.3397, "ab": 0.3333}, {"d": "2025-07-28", "acc": 0.2667, "n": 30, "lo": 0.109, "hi": 0.5195, "ab": 0.3333}, {"d": "2025-08-04", "acc": 0.1111, "n": 18, "lo": 0.0199, "hi": 0.435, "ab": 0.4444}, {"d": "2025-08-11", "acc": 0.0312, "n": 32, "lo": 0.0033, "hi": 0.2407, "ab": 0.4062}, {"d": "2025-08-18", "acc": 0.2632, "n": 38, "lo": 0.1181, "hi": 0.4879, "ab": 0.4211}, {"d": "2025-08-25", "acc": 0.1905, "n": 42, "lo": 0.0767, "hi": 0.4, "ab": 0.4762}, {"d": "2025-09-01", "acc": 0.0, "n": 20, "lo": 0.0, "hi": 0.2775, "ab": 0.5}, {"d": "2025-09-08", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.25}, {"d": "2025-09-15", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.2273}, {"d": "2025-09-22", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.3182}, {"d": "2025-09-29", "acc": 0.0, "n": 22, "lo": 0.0, "hi": 0.2588, "ab": 0.3636}, {"d": "2025-10-06", "acc": 0.2941, "n": 34, "lo": 0.1328, "hi": 0.5313, "ab": 0.2647}, {"d": "2025-10-13", "acc": 0.1111, "n": 36, "lo": 0.031, "hi": 0.328, "ab": 0.25}, {"d": "2025-10-20", "acc": 0.125, "n": 40, "lo": 0.0396, "hi": 0.3313, "ab": 0.45}, {"d": "2025-10-27", "acc": 0.1905, "n": 42, "lo": 0.0767, "hi": 0.4, "ab": 0.2619}, {"d": "2025-11-03", "acc": 0.3846, "n": 26, "lo": 0.1771, "hi": 0.6448, "ab": 0.1538}, {"d": "2025-11-10", "acc": 0.1071, "n": 28, "lo": 0.0251, "hi": 0.3583, "ab": 0.25}, {"d": "2025-11-17", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.2727}, {"d": "2025-11-24", "acc": 0.0667, "n": 30, "lo": 0.0119, "hi": 0.2982, "ab": 0.2}, {"d": "2025-12-01", "acc": 0.1667, "n": 30, "lo": 0.0531, "hi": 0.4161, "ab": 0.3667}, {"d": "2025-12-08", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.2727}, {"d": "2025-12-15", "acc": 0.0526, "n": 38, "lo": 0.0094, "hi": 0.2464, "ab": 0.3158}, {"d": "2025-12-22", "acc": 0.2273, "n": 22, "lo": 0.0733, "hi": 0.5225, "ab": 0.1818}, {"d": "2025-12-29", "acc": 0.0833, "n": 24, "lo": 0.0149, "hi": 0.3539, "ab": 0.375}, {"d": "2026-01-05", "acc": 0.0385, "n": 26, "lo": 0.004, "hi": 0.2835, "ab": 0.3846}, {"d": "2026-01-12", "acc": 0.2188, "n": 32, "lo": 0.0833, "hi": 0.4631, "ab": 0.4062}, {"d": "2026-01-19", "acc": 0.0882, "n": 34, "lo": 0.0206, "hi": 0.3076, "ab": 0.4118}, {"d": "2026-01-26", "acc": 0.0385, "n": 26, "lo": 0.004, "hi": 0.2835, "ab": 0.3846}, {"d": "2026-02-02", "acc": 0.125, "n": 32, "lo": 0.035, "hi": 0.3602, "ab": 0.4062}, {"d": "2026-02-09", "acc": 0.0667, "n": 30, "lo": 0.0119, "hi": 0.2982, "ab": 0.5333}, {"d": "2026-02-16", "acc": 0.1111, "n": 18, "lo": 0.0199, "hi": 0.435, "ab": 0.5}, {"d": "2026-02-23", "acc": 0.0, "n": 28, "lo": 0.0, "hi": 0.2153, "ab": 0.5714}, {"d": "2026-03-02", "acc": 0.2059, "n": 34, "lo": 0.0782, "hi": 0.442, "ab": 0.5}, {"d": "2026-03-09", "acc": 0.0714, "n": 42, "lo": 0.0167, "hi": 0.2587, "ab": 0.6429}, {"d": "2026-03-16", "acc": 0.1071, "n": 28, "lo": 0.0251, "hi": 0.3583, "ab": 0.5357}, {"d": "2026-03-23", "acc": 0.1, "n": 20, "lo": 0.0179, "hi": 0.4042, "ab": 0.55}, {"d": "2026-03-30", "acc": 0.125, "n": 32, "lo": 0.035, "hi": 0.3602, "ab": 0.6562}, {"d": "2026-04-06", "acc": 0.1471, "n": 34, "lo": 0.0467, "hi": 0.3775, "ab": 0.5}, {"d": "2026-04-13", "acc": 0.1176, "n": 34, "lo": 0.0329, "hi": 0.3434, "ab": 0.6176}, {"d": "2026-04-20", "acc": 0.0667, "n": 30, "lo": 0.0119, "hi": 0.2982, "ab": 0.7}, {"d": "2026-04-27", "acc": 0.0263, "n": 38, "lo": 0.0028, "hi": 0.2092, "ab": 0.6053}, {"d": "2026-05-04", "acc": 0.1176, "n": 34, "lo": 0.0329, "hi": 0.3434, "ab": 0.4706}, {"d": "2026-05-11", "acc": 0.1176, "n": 34, "lo": 0.0329, "hi": 0.3434, "ab": 0.3235}, {"d": "2026-05-18", "acc": 0.0769, "n": 26, "lo": 0.0137, "hi": 0.3331, "ab": 0.6923}, {"d": "2026-05-25", "acc": 0.0357, "n": 28, "lo": 0.0037, "hi": 0.2676, "ab": 0.7143}, {"d": "2026-06-01", "acc": 0.0357, "n": 28, "lo": 0.0037, "hi": 0.2676, "ab": 0.5357}, {"d": "2026-06-08", "acc": 0.1818, "n": 22, "lo": 0.0514, "hi": 0.477, "ab": 0.4091}, {"d": "2026-06-15", "acc": 0.0769, "n": 26, "lo": 0.0137, "hi": 0.3331, "ab": 0.6538}, {"d": "2026-06-22", "acc": 0.0, "n": 26, "lo": 0.0, "hi": 0.2281, "ab": 0.6923}, {"d": "2026-06-29", "acc": 0.0278, "n": 36, "lo": 0.0029, "hi": 0.2188, "ab": 0.4444}, {"d": "2026-07-06", "acc": 0.0588, "n": 34, "lo": 0.0105, "hi": 0.2698, "ab": 0.6765}, {"d": "2026-07-13", "acc": 0.0, "n": 38, "lo": 0.0, "hi": 0.1682, "ab": 0.7368}, {"d": "2026-07-20", "acc": 0.0, "n": 28, "lo": 0.0, "hi": 0.2153, "ab": 0.7143}, {"d": "2026-07-27", "acc": 0.125, "n": 16, "lo": 0.0224, "hi": 0.4709, "ab": 0.5625}], "samples": [{"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 85", "ok": true, "abstain": false, "conf": 85}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 85 The event described matches Turkey's blocking of Instagram in early", "ok": true, "abstain": false, "conf": 85}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, a South American government became the third country in the world to recognise Edmundo González as the rightful winner of Venezuela's disputed presidential election. Which country was it?", "correct_opt": "Argentina", "model_opt": "Costa Rica", "raw": "D, 45", "ok": false, "abstain": false, "conf": 45}, {"band": "mid", "date": "2025-01-02", "stem": "In early January 2025, a new governor was inaugurated in Puerto Rico. Who took the oath of office?", "correct_opt": "Jenniffer González Colón", "model_opt": "Jenniffer González Colón", "raw": "F, 85", "ok": true, "abstain": false, "conf": 85}, {"band": "mid", "date": "2025-01-02", "stem": "In early January 2025, Italy called in Tehran's ambassador and pressed for the release of an Italian journalist who had been detained in Iran. For which Italian newspaper did the reporter work?", "correct_opt": "Il Foglio", "model_opt": "Corriere della Sera", "raw": "B, 75", "ok": false, "abstain": false, "conf": 75}, {"band": "mid", "date": "2025-01-02", "stem": "In early January 2025, researchers announced that a working quarry in England had yielded the country's biggest known assemblage of tracks left by Middle Jurassic dinosaurs. In which English county was the site located?", "correct_opt": "Oxfordshire", "model_opt": "Oxfordshire", "raw": "D, 75", "ok": true, "abstain": false, "conf": 75}, {"band": "mid", "date": "2025-01-06", "stem": "In early January 2025, Venezuela cut off diplomatic ties with a South American country after that country's president publicly recognised Edmundo González as Venezuela's rightful president-elect. Which country was it?", "correct_opt": "Paraguay", "model_opt": "Ecuador", "raw": "D, 75", "ok": false, "abstain": false, "conf": 75}, {"band": "post", "date": "2025-05-09", "stem": "In mid-May 2025, Indian cruise missiles struck a major Pakistani air base in Punjab, with reports of multiple blasts and Pakistani air defences engaging. Which air base was hit?", "correct_opt": "Nur Khan", "model_opt": null, "raw": "I cannot provide information about this event because it appears to describe a military c…", "ok": false, "abstain": true, "conf": null}, {"band": "post", "date": "2025-05-09", "stem": "In mid-May 2025, a diver died while preparations were under way to raise a luxury yacht that had gone down off Sicily the previous year. What was the nationality of the diver who was killed?", "correct_opt": "Dutch", "model_opt": "Italian", "raw": "G, 45", "ok": false, "abstain": false, "conf": 45}, {"band": "post", "date": "2025-05-10", "stem": "In mid-May 2025, a Soviet spacecraft launched in 1972 that never escaped Earth orbit finally fell back through the atmosphere. According to the Russian space agency, where did it come down?", "correct_opt": "The Indian Ocean west of Indonesia", "model_opt": "The South Pacific east of New Zealand", "raw": "E, 45", "ok": false, "abstain": false, "conf": 45}]}, {"id": "claude-haiku-4-5-20251001", "name": "Claude Haiku 4.5", "provider": "anthropic", "family": "haiku", "release": "2025-10-15", "pub_rel": "2025-02", "pub_train": "2025-07", "pub_rel_d": "2025-02-28", "pub_train_d": "2025-07-31", "fit": {"midpoint": "2025-02-15", "width": 7.87, "decay_start": "2025-01-23", "floor_reach": "2025-03-28", "floor": 0.0298, "plateau": 0.3, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2025-02-03", "2025-02-26"]}, "weekly": [{"d": "2024-07-29", "acc": 0.375, "n": 16, "lo": 0.1368, "hi": 0.6943, "ab": 0.0625}, {"d": "2024-08-05", "acc": 0.1538, "n": 26, "lo": 0.0433, "hi": 0.4224, "ab": 0.1538}, {"d": "2024-08-12", "acc": 0.25, "n": 24, "lo": 0.0889, "hi": 0.5323, "ab": 0.125}, {"d": "2024-08-19", "acc": 0.2308, "n": 26, "lo": 0.0818, "hi": 0.5026, "ab": 0.1538}, {"d": "2024-08-26", "acc": 0.0, "n": 22, "lo": 0.0, "hi": 0.2588, "ab": 0.0455}, {"d": "2024-09-02", "acc": 0.2188, "n": 32, "lo": 0.0833, "hi": 0.4631, "ab": 0.125}, {"d": "2024-09-09", "acc": 0.1538, "n": 26, "lo": 0.0433, "hi": 0.4224, "ab": 0.1538}, {"d": "2024-09-16", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.1786}, {"d": "2024-09-23", "acc": 0.5909, "n": 22, "lo": 0.3161, "hi": 0.8186, "ab": 0.2273}, {"d": "2024-09-30", "acc": 0.1667, "n": 24, "lo": 0.047, "hi": 0.448, "ab": 0.25}, {"d": "2024-10-07", "acc": 0.2143, "n": 14, "lo": 0.0511, "hi": 0.5799, "ab": 0.2143}, {"d": "2024-10-14", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.1818}, {"d": "2024-10-21", "acc": 0.25, "n": 16, "lo": 0.0715, "hi": 0.5907, "ab": 0.1875}, {"d": "2024-10-28", "acc": 0.375, "n": 24, "lo": 0.165, "hi": 0.6456, "ab": 0.2083}, {"d": "2024-11-04", "acc": 0.2273, "n": 22, "lo": 0.0733, "hi": 0.5225, "ab": 0.3636}, {"d": "2024-11-11", "acc": 0.25, "n": 24, "lo": 0.0889, "hi": 0.5323, "ab": 0.0833}, {"d": "2024-11-18", "acc": 0.3333, "n": 18, "lo": 0.1206, "hi": 0.6458, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.3235, "n": 34, "lo": 0.1526, "hi": 0.5595, "ab": 0.2353}, {"d": "2024-12-02", "acc": 0.0455, "n": 22, "lo": 0.0048, "hi": 0.3215, "ab": 0.2727}, {"d": "2024-12-09", "acc": 0.3, "n": 30, "lo": 0.1299, "hi": 0.5517, "ab": 0.3}, {"d": "2024-12-16", "acc": 0.1579, "n": 38, "lo": 0.0552, "hi": 0.3757, "ab": 0.3684}, {"d": "2024-12-23", "acc": 0.1, "n": 40, "lo": 0.0279, "hi": 0.301, "ab": 0.375}, {"d": "2024-12-30", "acc": 0.2, "n": 10, "lo": 0.0362, "hi": 0.6245, "ab": 0.4}, {"d": "2025-01-06", "acc": 0.5, "n": 18, "lo": 0.2265, "hi": 0.7735, "ab": 0.0}, {"d": "2025-01-13", "acc": 0.1875, "n": 16, "lo": 0.0445, "hi": 0.5332, "ab": 0.125}, {"d": "2025-01-20", "acc": 0.5, "n": 22, "lo": 0.2456, "hi": 0.7544, "ab": 0.2727}, {"d": "2025-01-27", "acc": 0.4, "n": 30, "lo": 0.1982, "hi": 0.6425, "ab": 0.4}, {"d": "2025-02-03", "acc": 0.1429, "n": 28, "lo": 0.0401, "hi": 0.3994, "ab": 0.5714}, {"d": "2025-02-10", "acc": 0.0909, "n": 22, "lo": 0.0162, "hi": 0.3774, "ab": 0.2727}, {"d": "2025-02-17", "acc": 0.1, "n": 30, "lo": 0.0234, "hi": 0.3397, "ab": 0.6}, {"d": "2025-02-24", "acc": 0.1429, "n": 28, "lo": 0.0401, "hi": 0.3994, "ab": 0.6429}, {"d": "2025-03-03", "acc": 0.0312, "n": 32, "lo": 0.0033, "hi": 0.2407, "ab": 0.8438}, {"d": "2025-03-10", "acc": 0.0833, "n": 36, "lo": 0.0195, "hi": 0.2938, "ab": 0.8611}, {"d": "2025-03-17", "acc": 0.0, "n": 26, "lo": 0.0, "hi": 0.2281, "ab": 0.7308}, {"d": "2025-03-24", "acc": 0.0, "n": 20, "lo": 0.0, "hi": 0.2775, "ab": 0.95}, {"d": "2025-03-31", "acc": 0.0, "n": 28, "lo": 0.0, "hi": 0.2153, "ab": 0.75}, {"d": "2025-04-07", "acc": 0.0, "n": 36, "lo": 0.0, "hi": 0.1759, "ab": 0.8056}, {"d": "2025-04-14", "acc": 0.0, "n": 22, "lo": 0.0, "hi": 0.2588, "ab": 0.9091}, {"d": "2025-04-21", "acc": 0.0588, "n": 34, "lo": 0.0105, "hi": 0.2698, "ab": 0.8824}, {"d": "2025-04-28", "acc": 0.0, "n": 28, "lo": 0.0, "hi": 0.2153, "ab": 0.9286}, {"d": "2025-05-05", "acc": 0.0357, "n": 28, "lo": 0.0037, "hi": 0.2676, "ab": 0.8929}, {"d": "2025-05-12", "acc": 0.0, "n": 38, "lo": 0.0, "hi": 0.1682, "ab": 0.9474}, {"d": "2025-05-19", "acc": 0.1667, "n": 30, "lo": 0.0531, "hi": 0.4161, "ab": 0.8333}, {"d": "2025-05-26", "acc": 0.0, "n": 24, "lo": 0.0, "hi": 0.2425, "ab": 1.0}, {"d": "2025-06-02", "acc": 0.0588, "n": 34, "lo": 0.0105, "hi": 0.2698, "ab": 0.8235}, {"d": "2025-06-09", "acc": 0.0, "n": 24, "lo": 0.0, "hi": 0.2425, "ab": 0.9583}, {"d": "2025-06-16", "acc": 0.0, "n": 30, "lo": 0.0, "hi": 0.2039, "ab": 0.9667}, {"d": "2025-06-23", "acc": 0.0588, "n": 34, "lo": 0.0105, "hi": 0.2698, "ab": 0.8235}, {"d": "2025-06-30", "acc": 0.0, "n": 38, "lo": 0.0, "hi": 0.1682, "ab": 0.9474}, {"d": "2025-07-07", "acc": 0.0, "n": 24, "lo": 0.0, "hi": 0.2425, "ab": 0.9167}, {"d": "2025-07-14", "acc": 0.0, "n": 22, "lo": 0.0, "hi": 0.2588, "ab": 0.6818}, {"d": "2025-07-21", "acc": 0.0, "n": 30, "lo": 0.0, "hi": 0.2039, "ab": 0.9667}, {"d": "2025-07-28", "acc": 0.0, "n": 30, "lo": 0.0, "hi": 0.2039, "ab": 0.9667}, {"d": "2025-08-04", "acc": 0.0, "n": 18, "lo": 0.0, "hi": 0.2992, "ab": 0.8889}, {"d": "2025-08-11", "acc": 0.0, "n": 32, "lo": 0.0, "hi": 0.1936, "ab": 0.7188}, {"d": "2025-08-18", "acc": 0.0263, "n": 38, "lo": 0.0028, "hi": 0.2092, "ab": 0.8158}, {"d": "2025-08-25", "acc": 0.0238, "n": 42, "lo": 0.0025, "hi": 0.1924, "ab": 0.6667}, {"d": "2025-09-01", "acc": 0.05, "n": 20, "lo": 0.0052, "hi": 0.3445, "ab": 0.8}, {"d": "2025-09-08", "acc": 0.1071, "n": 28, "lo": 0.0251, "hi": 0.3583, "ab": 0.7857}, {"d": "2025-09-15", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.7273}, {"d": "2025-09-22", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.7273}, {"d": "2025-09-29", "acc": 0.0, "n": 22, "lo": 0.0, "hi": 0.2588, "ab": 0.9091}, {"d": "2025-10-06", "acc": 0.0588, "n": 34, "lo": 0.0105, "hi": 0.2698, "ab": 0.8529}, {"d": "2025-10-13", "acc": 0.0, "n": 36, "lo": 0.0, "hi": 0.1759, "ab": 0.75}, {"d": "2025-10-20", "acc": 0.025, "n": 40, "lo": 0.0026, "hi": 0.2005, "ab": 0.875}, {"d": "2025-10-27", "acc": 0.119, "n": 42, "lo": 0.0376, "hi": 0.3183, "ab": 0.7857}, {"d": "2025-11-03", "acc": 0.1538, "n": 26, "lo": 0.0433, "hi": 0.4224, "ab": 0.5769}, {"d": "2025-11-10", "acc": 0.0714, "n": 28, "lo": 0.0127, "hi": 0.3147, "ab": 0.7143}, {"d": "2025-11-17", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.6818}, {"d": "2025-11-24", "acc": 0.0333, "n": 30, "lo": 0.0035, "hi": 0.2535, "ab": 0.9}, {"d": "2025-12-01", "acc": 0.0333, "n": 30, "lo": 0.0035, "hi": 0.2535, "ab": 0.9}, {"d": "2025-12-08", "acc": 0.0, "n": 22, "lo": 0.0, "hi": 0.2588, "ab": 0.7727}, {"d": "2025-12-15", "acc": 0.0, "n": 38, "lo": 0.0, "hi": 0.1682, "ab": 0.8421}, {"d": "2025-12-22", "acc": 0.0, "n": 22, "lo": 0.0, "hi": 0.2588, "ab": 0.7727}, {"d": "2025-12-29", "acc": 0.0417, "n": 24, "lo": 0.0044, "hi": 0.3013, "ab": 0.7917}, {"d": "2026-01-05", "acc": 0.0769, "n": 26, "lo": 0.0137, "hi": 0.3331, "ab": 0.7308}, {"d": "2026-01-12", "acc": 0.0938, "n": 32, "lo": 0.022, "hi": 0.3229, "ab": 0.6562}, {"d": "2026-01-19", "acc": 0.0588, "n": 34, "lo": 0.0105, "hi": 0.2698, "ab": 0.8235}, {"d": "2026-01-26", "acc": 0.0, "n": 26, "lo": 0.0, "hi": 0.2281, "ab": 0.8462}, {"d": "2026-02-02", "acc": 0.0, "n": 32, "lo": 0.0, "hi": 0.1936, "ab": 0.9375}, {"d": "2026-02-09", "acc": 0.0, "n": 30, "lo": 0.0, "hi": 0.2039, "ab": 0.9333}, {"d": "2026-02-16", "acc": 0.0, "n": 18, "lo": 0.0, "hi": 0.2992, "ab": 0.8889}, {"d": "2026-02-23", "acc": 0.0357, "n": 28, "lo": 0.0037, "hi": 0.2676, "ab": 0.9286}, {"d": "2026-03-02", "acc": 0.0294, "n": 34, "lo": 0.0031, "hi": 0.2292, "ab": 0.8529}, {"d": "2026-03-09", "acc": 0.0238, "n": 42, "lo": 0.0025, "hi": 0.1924, "ab": 0.9048}, {"d": "2026-03-16", "acc": 0.0714, "n": 28, "lo": 0.0127, "hi": 0.3147, "ab": 0.8571}, {"d": "2026-03-23", "acc": 0.0, "n": 20, "lo": 0.0, "hi": 0.2775, "ab": 0.95}, {"d": "2026-03-30", "acc": 0.0, "n": 32, "lo": 0.0, "hi": 0.1936, "ab": 0.9375}, {"d": "2026-04-06", "acc": 0.0588, "n": 34, "lo": 0.0105, "hi": 0.2698, "ab": 0.9412}, {"d": "2026-04-13", "acc": 0.0, "n": 34, "lo": 0.0, "hi": 0.1843, "ab": 0.9118}, {"d": "2026-04-20", "acc": 0.0, "n": 30, "lo": 0.0, "hi": 0.2039, "ab": 0.9333}, {"d": "2026-04-27", "acc": 0.0526, "n": 38, "lo": 0.0094, "hi": 0.2464, "ab": 0.8421}, {"d": "2026-05-04", "acc": 0.0588, "n": 34, "lo": 0.0105, "hi": 0.2698, "ab": 0.9118}, {"d": "2026-05-11", "acc": 0.0, "n": 34, "lo": 0.0, "hi": 0.1843, "ab": 0.8235}, {"d": "2026-05-18", "acc": 0.0, "n": 26, "lo": 0.0, "hi": 0.2281, "ab": 1.0}, {"d": "2026-05-25", "acc": 0.0, "n": 28, "lo": 0.0, "hi": 0.2153, "ab": 0.8214}, {"d": "2026-06-01", "acc": 0.0, "n": 28, "lo": 0.0, "hi": 0.2153, "ab": 1.0}, {"d": "2026-06-08", "acc": 0.0, "n": 22, "lo": 0.0, "hi": 0.2588, "ab": 0.7727}, {"d": "2026-06-15", "acc": 0.0, "n": 26, "lo": 0.0, "hi": 0.2281, "ab": 0.9615}, {"d": "2026-06-22", "acc": 0.0, "n": 26, "lo": 0.0, "hi": 0.2281, "ab": 0.9615}, {"d": "2026-06-29", "acc": 0.0, "n": 36, "lo": 0.0, "hi": 0.1759, "ab": 0.9167}, {"d": "2026-07-06", "acc": 0.0, "n": 34, "lo": 0.0, "hi": 0.1843, "ab": 0.9118}, {"d": "2026-07-13", "acc": 0.0, "n": 38, "lo": 0.0, "hi": 0.1682, "ab": 0.9474}, {"d": "2026-07-20", "acc": 0.0, "n": 28, "lo": 0.0, "hi": 0.2153, "ab": 0.9286}, {"d": "2026-07-27", "acc": 0.0, "n": 16, "lo": 0.0, "hi": 0.3244, "ab": 0.9375}], "samples": [{"band": "pre", "date": "2024-08-03", "stem": "In early August 2024, Vietnam's ruling party formally installed a new General Secretary after the death of Nguyễn Phú Trọng the previous month. Who took the post?", "correct_opt": "Tô Lâm", "model_opt": "Tô Lâm", "raw": "C, 92 Tô Lâm was elected as the", "ok": true, "abstain": false, "conf": 92}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, two U.S. federal agencies jointly sued TikTok and ByteDance in court. What was the alleged violation at the heart of the complaint?", "correct_opt": "Collecting and using personal data from children under 13", "model_opt": "Collecting and using personal data from children under 13", "raw": "F, 75 In August 2024, the FTC and Department", "ok": true, "abstain": false, "conf": 75}, {"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Switzerland", "raw": "G, 85 Switzerland acted as the mediator for the major prisoner exchange", "ok": false, "abstain": false, "conf": 85}, {"band": "mid", "date": "2025-01-24", "stem": "In late January 2025, Oscar nominations were announced, and for the first time a Brazilian production was shortlisted for Best Picture. Which film was it?", "correct_opt": "I'm Still Here", "model_opt": "I'm Still Here", "raw": "B, 92 The film \"I'm Still Here\" (Ainda", "ok": true, "abstain": false, "conf": 92}, {"band": "mid", "date": "2025-01-25", "stem": "In late January 2025, a first-time Grand Slam champion was crowned in the women's singles at the Australian Open, beating the two-time defending champion in three sets. Who won the title?", "correct_opt": "Madison Keys", "model_opt": null, "raw": "I need to determine who won the 2025 Australian Open women's singles", "ok": false, "abstain": true, "conf": null}, {"band": "mid", "date": "2025-01-24", "stem": "In late January 2025, a nominee for U.S. Secretary of Defense was confirmed only after the vice president broke a tie in the Senate. Who was confirmed?", "correct_opt": "Pete Hegseth", "model_opt": "Pete Hegseth", "raw": "A, 85 Pete Hegseth was nominated for Secretary of Defense", "ok": true, "abstain": false, "conf": 85}, {"band": "mid", "date": "2025-01-27", "stem": "In late January 2025, after talks in Beijing between India's foreign secretary and China's foreign minister, the two countries agreed to restore something suspended for about five years. What did they agree to restore?", "correct_opt": "Direct passenger flights between the two countries", "model_opt": "Annual defence ministry hotline talks", "raw": "G, 75", "ok": false, "abstain": false, "conf": 75}, {"band": "post", "date": "2025-03-30", "stem": "In late March 2025, a privately developed orbital rocket lifted off from a European mainland spaceport for the first time, but fell back and exploded after roughly thirty seconds of flight. From which site was it launch…", "correct_opt": "Andøya, Norway", "model_opt": null, "raw": "I need to identify which European mainland spaceport had a private orbital rocket launch", "ok": false, "abstain": true, "conf": null}, {"band": "post", "date": "2025-03-31", "stem": "In late March 2025, Brazilian police stopped a diplomatic vehicle during a highway check in Ponta Porã and found hundreds of smuggled electronics and luxury items worth some US$700,000. The detained man worked for the c…", "correct_opt": "Syria", "model_opt": null, "raw": "I don't have reliable information about a specific incident in late March 2025", "ok": false, "abstain": true, "conf": null}, {"band": "post", "date": "2025-04-01", "stem": "In early April 2025, an explosion tore through an unlicensed fireworks workshop in India's Gujarat state, killing more than twenty workers. In which town did the blast occur?", "correct_opt": "Deesa", "model_opt": null, "raw": "I don't have reliable information about a specific fireworks workshop explosion in Gujara…", "ok": false, "abstain": true, "conf": null}]}, {"id": "claude-opus-4-5-20251101", "name": "Claude Opus 4.5", "provider": "anthropic", "family": "opus", "release": "2025-11-24", "pub_rel": "2025-05", "pub_train": "2025-08", "pub_rel_d": "2025-05-31", "pub_train_d": "2025-08-31", "fit": {"midpoint": "2025-03-23", "width": 61.99, "decay_start": "2024-10-10", "floor_reach": "2025-12-02", "floor": 0.1682, "plateau": 0.6839, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2025-02-20", "2025-04-16"]}, "weekly": [{"d": "2024-07-29", "acc": 0.75, "n": 16, "lo": 0.4093, "hi": 0.9285, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.5769, "n": 26, "lo": 0.3228, "hi": 0.796, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.625, "n": 24, "lo": 0.3544, "hi": 0.835, "ab": 0.0}, {"d": "2024-08-19", "acc": 0.6538, "n": 26, "lo": 0.3888, "hi": 0.8487, "ab": 0.0769}, {"d": "2024-08-26", "acc": 0.5, "n": 22, "lo": 0.2456, "hi": 0.7544, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.8125, "n": 32, "lo": 0.5699, "hi": 0.9341, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.5385, "n": 26, "lo": 0.2914, "hi": 0.7679, "ab": 0.0}, {"d": "2024-09-16", "acc": 0.6786, "n": 28, "lo": 0.42, "hi": 0.8602, "ab": 0.0}, {"d": "2024-09-23", "acc": 0.7727, "n": 22, "lo": 0.4775, "hi": 0.9267, "ab": 0.0}, {"d": "2024-09-30", "acc": 0.875, "n": 24, "lo": 0.5975, "hi": 0.9706, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.5714, "n": 14, "lo": 0.2505, "hi": 0.8418, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.5909, "n": 22, "lo": 0.3161, "hi": 0.8186, "ab": 0.0}, {"d": "2024-10-21", "acc": 0.5625, "n": 16, "lo": 0.2589, "hi": 0.8255, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.375, "n": 24, "lo": 0.165, "hi": 0.6456, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.5909, "n": 22, "lo": 0.3161, "hi": 0.8186, "ab": 0.0}, {"d": "2024-11-11", "acc": 0.7083, "n": 24, "lo": 0.4284, "hi": 0.8873, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.6667, "n": 18, "lo": 0.3542, "hi": 0.8794, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.6471, "n": 34, "lo": 0.413, "hi": 0.8269, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.3636, "n": 22, "lo": 0.1517, "hi": 0.6462, "ab": 0.0}, {"d": "2024-12-09", "acc": 0.8667, "n": 30, "lo": 0.6212, "hi": 0.9626, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.6316, "n": 38, "lo": 0.4104, "hi": 0.8085, "ab": 0.0}, {"d": "2024-12-23", "acc": 0.425, "n": 40, "lo": 0.2383, "hi": 0.6359, "ab": 0.0}, {"d": "2024-12-30", "acc": 0.8, "n": 10, "lo": 0.3755, "hi": 0.9638, "ab": 0.0}, {"d": "2025-01-06", "acc": 0.8333, "n": 18, "lo": 0.5067, "hi": 0.9605, "ab": 0.0}, {"d": "2025-01-13", "acc": 0.75, "n": 16, "lo": 0.4093, "hi": 0.9285, "ab": 0.0}, {"d": "2025-01-20", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2025-01-27", "acc": 0.6, "n": 30, "lo": 0.3575, "hi": 0.8018, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.0}, {"d": "2025-02-10", "acc": 0.5, "n": 22, "lo": 0.2456, "hi": 0.7544, "ab": 0.0}, {"d": "2025-02-17", "acc": 0.5333, "n": 30, "lo": 0.3012, "hi": 0.7519, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.3214, "n": 28, "lo": 0.1398, "hi": 0.58, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.3125, "n": 32, "lo": 0.1416, "hi": 0.556, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.4167, "n": 36, "lo": 0.224, "hi": 0.6386, "ab": 0.0}, {"d": "2025-03-17", "acc": 0.4615, "n": 26, "lo": 0.2321, "hi": 0.7086, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.4, "n": 20, "lo": 0.1682, "hi": 0.6873, "ab": 0.05}, {"d": "2025-03-31", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.0357}, {"d": "2025-04-07", "acc": 0.3611, "n": 36, "lo": 0.1826, "hi": 0.5885, "ab": 0.0278}, {"d": "2025-04-14", "acc": 0.2273, "n": 22, "lo": 0.0733, "hi": 0.5225, "ab": 0.0455}, {"d": "2025-04-21", "acc": 0.2059, "n": 34, "lo": 0.0782, "hi": 0.442, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.5, "n": 28, "lo": 0.268, "hi": 0.732, "ab": 0.0}, {"d": "2025-05-05", "acc": 0.6786, "n": 28, "lo": 0.42, "hi": 0.8602, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.3684, "n": 38, "lo": 0.1915, "hi": 0.5896, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.4333, "n": 30, "lo": 0.2228, "hi": 0.6711, "ab": 0.0}, {"d": "2025-05-26", "acc": 0.2917, "n": 24, "lo": 0.1127, "hi": 0.5716, "ab": 0.0833}, {"d": "2025-06-02", "acc": 0.1471, "n": 34, "lo": 0.0467, "hi": 0.3775, "ab": 0.0}, {"d": "2025-06-09", "acc": 0.3333, "n": 24, "lo": 0.1381, "hi": 0.6094, "ab": 0.0833}, {"d": "2025-06-16", "acc": 0.2667, "n": 30, "lo": 0.109, "hi": 0.5195, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.1471, "n": 34, "lo": 0.0467, "hi": 0.3775, "ab": 0.0588}, {"d": "2025-06-30", "acc": 0.3158, "n": 38, "lo": 0.1536, "hi": 0.5399, "ab": 0.1316}, {"d": "2025-07-07", "acc": 0.1667, "n": 24, "lo": 0.047, "hi": 0.448, "ab": 0.1667}, {"d": "2025-07-14", "acc": 0.1818, "n": 22, "lo": 0.0514, "hi": 0.477, "ab": 0.2273}, {"d": "2025-07-21", "acc": 0.2667, "n": 30, "lo": 0.109, "hi": 0.5195, "ab": 0.0667}, {"d": "2025-07-28", "acc": 0.1333, "n": 30, "lo": 0.0374, "hi": 0.3788, "ab": 0.2}, {"d": "2025-08-04", "acc": 0.1111, "n": 18, "lo": 0.0199, "hi": 0.435, "ab": 0.2778}, {"d": "2025-08-11", "acc": 0.0312, "n": 32, "lo": 0.0033, "hi": 0.2407, "ab": 0.0938}, {"d": "2025-08-18", "acc": 0.2368, "n": 38, "lo": 0.1012, "hi": 0.461, "ab": 0.1316}, {"d": "2025-08-25", "acc": 0.2857, "n": 42, "lo": 0.1381, "hi": 0.4996, "ab": 0.1429}, {"d": "2025-09-01", "acc": 0.25, "n": 20, "lo": 0.0809, "hi": 0.5578, "ab": 0.3}, {"d": "2025-09-08", "acc": 0.3214, "n": 28, "lo": 0.1398, "hi": 0.58, "ab": 0.25}, {"d": "2025-09-15", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.1818}, {"d": "2025-09-22", "acc": 0.2273, "n": 22, "lo": 0.0733, "hi": 0.5225, "ab": 0.2273}, {"d": "2025-09-29", "acc": 0.0455, "n": 22, "lo": 0.0048, "hi": 0.3215, "ab": 0.1818}, {"d": "2025-10-06", "acc": 0.3235, "n": 34, "lo": 0.1526, "hi": 0.5595, "ab": 0.0588}, {"d": "2025-10-13", "acc": 0.25, "n": 36, "lo": 0.1071, "hi": 0.4808, "ab": 0.1389}, {"d": "2025-10-20", "acc": 0.175, "n": 40, "lo": 0.0661, "hi": 0.3886, "ab": 0.225}, {"d": "2025-10-27", "acc": 0.1905, "n": 42, "lo": 0.0767, "hi": 0.4, "ab": 0.1429}, {"d": "2025-11-03", "acc": 0.3846, "n": 26, "lo": 0.1771, "hi": 0.6448, "ab": 0.0}, {"d": "2025-11-10", "acc": 0.1071, "n": 28, "lo": 0.0251, "hi": 0.3583, "ab": 0.0714}, {"d": "2025-11-17", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.1818}, {"d": "2025-11-24", "acc": 0.3, "n": 30, "lo": 0.1299, "hi": 0.5517, "ab": 0.2}, {"d": "2025-12-01", "acc": 0.1667, "n": 30, "lo": 0.0531, "hi": 0.4161, "ab": 0.0333}, {"d": "2025-12-08", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.1053, "n": 38, "lo": 0.0294, "hi": 0.3139, "ab": 0.1579}, {"d": "2025-12-22", "acc": 0.0455, "n": 22, "lo": 0.0048, "hi": 0.3215, "ab": 0.1818}, {"d": "2025-12-29", "acc": 0.0417, "n": 24, "lo": 0.0044, "hi": 0.3013, "ab": 0.1667}, {"d": "2026-01-05", "acc": 0.0769, "n": 26, "lo": 0.0137, "hi": 0.3331, "ab": 0.1538}, {"d": "2026-01-12", "acc": 0.3125, "n": 32, "lo": 0.1416, "hi": 0.556, "ab": 0.1875}, {"d": "2026-01-19", "acc": 0.2059, "n": 34, "lo": 0.0782, "hi": 0.442, "ab": 0.1176}, {"d": "2026-01-26", "acc": 0.2308, "n": 26, "lo": 0.0818, "hi": 0.5026, "ab": 0.1154}, {"d": "2026-02-02", "acc": 0.2188, "n": 32, "lo": 0.0833, "hi": 0.4631, "ab": 0.0938}, {"d": "2026-02-09", "acc": 0.1667, "n": 30, "lo": 0.0531, "hi": 0.4161, "ab": 0.1667}, {"d": "2026-02-16", "acc": 0.1111, "n": 18, "lo": 0.0199, "hi": 0.435, "ab": 0.3333}, {"d": "2026-02-23", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.1071}, {"d": "2026-03-02", "acc": 0.2941, "n": 34, "lo": 0.1328, "hi": 0.5313, "ab": 0.2353}, {"d": "2026-03-09", "acc": 0.0714, "n": 42, "lo": 0.0167, "hi": 0.2587, "ab": 0.3095}, {"d": "2026-03-16", "acc": 0.1071, "n": 28, "lo": 0.0251, "hi": 0.3583, "ab": 0.25}, {"d": "2026-03-23", "acc": 0.2, "n": 20, "lo": 0.0567, "hi": 0.5098, "ab": 0.4}, {"d": "2026-03-30", "acc": 0.25, "n": 32, "lo": 0.1018, "hi": 0.495, "ab": 0.2812}, {"d": "2026-04-06", "acc": 0.2353, "n": 34, "lo": 0.0955, "hi": 0.4726, "ab": 0.2647}, {"d": "2026-04-13", "acc": 0.1471, "n": 34, "lo": 0.0467, "hi": 0.3775, "ab": 0.3235}, {"d": "2026-04-20", "acc": 0.1, "n": 30, "lo": 0.0234, "hi": 0.3397, "ab": 0.3333}, {"d": "2026-04-27", "acc": 0.0789, "n": 38, "lo": 0.0184, "hi": 0.2811, "ab": 0.3421}, {"d": "2026-05-04", "acc": 0.2353, "n": 34, "lo": 0.0955, "hi": 0.4726, "ab": 0.3824}, {"d": "2026-05-11", "acc": 0.0882, "n": 34, "lo": 0.0206, "hi": 0.3076, "ab": 0.2941}, {"d": "2026-05-18", "acc": 0.1923, "n": 26, "lo": 0.0616, "hi": 0.4634, "ab": 0.4231}, {"d": "2026-05-25", "acc": 0.1429, "n": 28, "lo": 0.0401, "hi": 0.3994, "ab": 0.5357}, {"d": "2026-06-01", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.3929}, {"d": "2026-06-08", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.3636}, {"d": "2026-06-15", "acc": 0.1538, "n": 26, "lo": 0.0433, "hi": 0.4224, "ab": 0.3846}, {"d": "2026-06-22", "acc": 0.1538, "n": 26, "lo": 0.0433, "hi": 0.4224, "ab": 0.5}, {"d": "2026-06-29", "acc": 0.1389, "n": 36, "lo": 0.0441, "hi": 0.3607, "ab": 0.2222}, {"d": "2026-07-06", "acc": 0.0, "n": 34, "lo": 0.0, "hi": 0.1843, "ab": 0.6176}, {"d": "2026-07-13", "acc": 0.1316, "n": 38, "lo": 0.0417, "hi": 0.3454, "ab": 0.5263}, {"d": "2026-07-20", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.3929}, {"d": "2026-07-27", "acc": 0.125, "n": 16, "lo": 0.0224, "hi": 0.4709, "ab": 0.4375}], "samples": [{"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 85", "ok": true, "abstain": false, "conf": 85}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 85", "ok": true, "abstain": false, "conf": 85}, {"band": "pre", "date": "2024-08-04", "stem": "In early August 2024, Algeria claimed its first Olympic gymnastics title in Paris. Which apparatus final produced that gold medal?", "correct_opt": "Women's uneven bars", "model_opt": "Men's parallel bars", "raw": "G, 45", "ok": false, "abstain": false, "conf": 45}, {"band": "mid", "date": "2024-10-13", "stem": "In mid-October 2024, an election committee in a Chinese special administrative region picked a new chief executive to succeed Ho Iat Seng. Who was chosen?", "correct_opt": "Sam Hou Fai", "model_opt": "Sam Hou Fai", "raw": "A, 92", "ok": true, "abstain": false, "conf": 92}, {"band": "mid", "date": "2024-10-14", "stem": "In mid-October 2024, the Israeli military said it had killed a Hezbollah commander who ran the anti-tank missile unit of the group's elite Redwan Force in a strike in southern Lebanon. Whom did it name?", "correct_opt": "Muhammad Kamel Naim", "model_opt": "Hassan Ali Sabra", "raw": "A, 45", "ok": false, "abstain": false, "conf": 45}, {"band": "mid", "date": "2024-10-14", "stem": "In mid-October 2024, a bus carrying university students overturned in Egypt's Suez Governorate, killing twelve people and injuring dozens more. Which university did the students attend?", "correct_opt": "Galala University", "model_opt": "Galala University", "raw": "G, 62", "ok": true, "abstain": false, "conf": 62}, {"band": "mid", "date": "2024-10-16", "stem": "In mid-October 2024, a bridge in Mississippi gave way into the river below while crews were preparing it for demolition, killing three workers. Into which river did the structure fall?", "correct_opt": "Strong River", "model_opt": "Big Black River", "raw": "A, 45", "ok": false, "abstain": false, "conf": 45}, {"band": "post", "date": "2025-12-02", "stem": "In early December 2025, Nigeria's president named a retired army general as the new defence minister to replace the outgoing minister who had stepped down. Who was appointed?", "correct_opt": "Christopher Gwabin Musa", "model_opt": "Lucky Irabor", "raw": "B, 45", "ok": false, "abstain": false, "conf": 45}, {"band": "post", "date": "2025-12-02", "stem": "In early December 2025, authorities in a South American country detained 17 children and nine adults belonging to an extremist Jewish sect over child-welfare concerns, then flew the whole group to New York. In which cou…", "correct_opt": "Colombia", "model_opt": "Panama", "raw": "C, 62", "ok": false, "abstain": false, "conf": 62}, {"band": "post", "date": "2025-12-02", "stem": "In early December 2025, an aid-organisation Cessna 208 Caravan was hijacked shortly after leaving Juba, South Sudan, and the gunman was overpowered after the plane stopped to refuel. Where did the hijacker demand to be …", "correct_opt": "Chad", "model_opt": "Egypt", "raw": "B, 45", "ok": false, "abstain": false, "conf": 45}]}, {"id": "claude-opus-4-6", "name": "Claude Opus 4.6", "provider": "anthropic", "family": "opus", "release": "2026-02-05", "pub_rel": "2025-05", "pub_train": "2025-08", "pub_rel_d": "2025-05-31", "pub_train_d": "2025-08-31", "fit": {"midpoint": "2025-03-24", "width": 44.99, "decay_start": "2024-12-02", "floor_reach": "2025-09-02", "floor": 0.2417, "plateau": 0.6953, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2025-02-28", "2025-04-16"]}, "weekly": [{"d": "2024-07-29", "acc": 0.6875, "n": 16, "lo": 0.3558, "hi": 0.8976, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.6538, "n": 26, "lo": 0.3888, "hi": 0.8487, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.6667, "n": 24, "lo": 0.3906, "hi": 0.8619, "ab": 0.0}, {"d": "2024-08-19", "acc": 0.7692, "n": 26, "lo": 0.4974, "hi": 0.9182, "ab": 0.0}, {"d": "2024-08-26", "acc": 0.6818, "n": 22, "lo": 0.3932, "hi": 0.8764, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.6875, "n": 32, "lo": 0.444, "hi": 0.8584, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.5385, "n": 26, "lo": 0.2914, "hi": 0.7679, "ab": 0.0}, {"d": "2024-09-16", "acc": 0.5357, "n": 28, "lo": 0.2965, "hi": 0.7596, "ab": 0.0}, {"d": "2024-09-23", "acc": 0.7727, "n": 22, "lo": 0.4775, "hi": 0.9267, "ab": 0.0}, {"d": "2024-09-30", "acc": 0.875, "n": 24, "lo": 0.5975, "hi": 0.9706, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.7143, "n": 14, "lo": 0.3589, "hi": 0.9178, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.6364, "n": 22, "lo": 0.3538, "hi": 0.8483, "ab": 0.0}, {"d": "2024-10-21", "acc": 0.625, "n": 16, "lo": 0.3057, "hi": 0.8632, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.375, "n": 24, "lo": 0.165, "hi": 0.6456, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.6818, "n": 22, "lo": 0.3932, "hi": 0.8764, "ab": 0.0}, {"d": "2024-11-11", "acc": 0.875, "n": 24, "lo": 0.5975, "hi": 0.9706, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.7778, "n": 18, "lo": 0.4526, "hi": 0.9368, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.7059, "n": 34, "lo": 0.4687, "hi": 0.8672, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.3636, "n": 22, "lo": 0.1517, "hi": 0.6462, "ab": 0.0}, {"d": "2024-12-09", "acc": 0.8667, "n": 30, "lo": 0.6212, "hi": 0.9626, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.6842, "n": 38, "lo": 0.4601, "hi": 0.8464, "ab": 0.0}, {"d": "2024-12-23", "acc": 0.525, "n": 40, "lo": 0.3205, "hi": 0.7215, "ab": 0.0}, {"d": "2024-12-30", "acc": 0.8, "n": 10, "lo": 0.3755, "hi": 0.9638, "ab": 0.0}, {"d": "2025-01-06", "acc": 0.8889, "n": 18, "lo": 0.565, "hi": 0.9801, "ab": 0.0}, {"d": "2025-01-13", "acc": 0.625, "n": 16, "lo": 0.3057, "hi": 0.8632, "ab": 0.0}, {"d": "2025-01-20", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2025-01-27", "acc": 0.6333, "n": 30, "lo": 0.3869, "hi": 0.8254, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.0}, {"d": "2025-02-10", "acc": 0.7273, "n": 22, "lo": 0.4343, "hi": 0.9025, "ab": 0.0}, {"d": "2025-02-17", "acc": 0.6333, "n": 30, "lo": 0.3869, "hi": 0.8254, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.4286, "n": 28, "lo": 0.2138, "hi": 0.6741, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.4375, "n": 32, "lo": 0.231, "hi": 0.6682, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.4167, "n": 36, "lo": 0.224, "hi": 0.6386, "ab": 0.0}, {"d": "2025-03-17", "acc": 0.5, "n": 26, "lo": 0.2612, "hi": 0.7388, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.2, "n": 20, "lo": 0.0567, "hi": 0.5098, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.5357, "n": 28, "lo": 0.2965, "hi": 0.7596, "ab": 0.0}, {"d": "2025-04-07", "acc": 0.5, "n": 36, "lo": 0.2903, "hi": 0.7097, "ab": 0.0}, {"d": "2025-04-14", "acc": 0.0909, "n": 22, "lo": 0.0162, "hi": 0.3774, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.2353, "n": 34, "lo": 0.0955, "hi": 0.4726, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.4643, "n": 28, "lo": 0.2404, "hi": 0.7035, "ab": 0.0}, {"d": "2025-05-05", "acc": 0.6429, "n": 28, "lo": 0.3876, "hi": 0.8366, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.3421, "n": 38, "lo": 0.1723, "hi": 0.565, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.4333, "n": 30, "lo": 0.2228, "hi": 0.6711, "ab": 0.0}, {"d": "2025-05-26", "acc": 0.4167, "n": 24, "lo": 0.1933, "hi": 0.6805, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.1765, "n": 34, "lo": 0.0619, "hi": 0.4103, "ab": 0.0}, {"d": "2025-06-09", "acc": 0.25, "n": 24, "lo": 0.0889, "hi": 0.5323, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.3667, "n": 30, "lo": 0.1746, "hi": 0.6131, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.2647, "n": 34, "lo": 0.1138, "hi": 0.5024, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.3684, "n": 38, "lo": 0.1915, "hi": 0.5896, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.2917, "n": 24, "lo": 0.1127, "hi": 0.5716, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.1818, "n": 22, "lo": 0.0514, "hi": 0.477, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.2667, "n": 30, "lo": 0.109, "hi": 0.5195, "ab": 0.0}, {"d": "2025-07-28", "acc": 0.2667, "n": 30, "lo": 0.109, "hi": 0.5195, "ab": 0.0}, {"d": "2025-08-04", "acc": 0.1667, "n": 18, "lo": 0.0395, "hi": 0.4933, "ab": 0.0}, {"d": "2025-08-11", "acc": 0.125, "n": 32, "lo": 0.035, "hi": 0.3602, "ab": 0.0}, {"d": "2025-08-18", "acc": 0.2632, "n": 38, "lo": 0.1181, "hi": 0.4879, "ab": 0.0}, {"d": "2025-08-25", "acc": 0.3095, "n": 42, "lo": 0.1548, "hi": 0.5231, "ab": 0.0}, {"d": "2025-09-01", "acc": 0.35, "n": 20, "lo": 0.1369, "hi": 0.6463, "ab": 0.0}, {"d": "2025-09-08", "acc": 0.5357, "n": 28, "lo": 0.2965, "hi": 0.7596, "ab": 0.0}, {"d": "2025-09-15", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.0}, {"d": "2025-09-22", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.0}, {"d": "2025-09-29", "acc": 0.3636, "n": 22, "lo": 0.1517, "hi": 0.6462, "ab": 0.0}, {"d": "2025-10-06", "acc": 0.2353, "n": 34, "lo": 0.0955, "hi": 0.4726, "ab": 0.0}, {"d": "2025-10-13", "acc": 0.3611, "n": 36, "lo": 0.1826, "hi": 0.5885, "ab": 0.0}, {"d": "2025-10-20", "acc": 0.2, "n": 40, "lo": 0.0807, "hi": 0.416, "ab": 0.025}, {"d": "2025-10-27", "acc": 0.2619, "n": 42, "lo": 0.1219, "hi": 0.4755, "ab": 0.0}, {"d": "2025-11-03", "acc": 0.4615, "n": 26, "lo": 0.2321, "hi": 0.7086, "ab": 0.0}, {"d": "2025-11-10", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.0}, {"d": "2025-11-17", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.4333, "n": 30, "lo": 0.2228, "hi": 0.6711, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.1333, "n": 30, "lo": 0.0374, "hi": 0.3788, "ab": 0.0}, {"d": "2025-12-08", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.1316, "n": 38, "lo": 0.0417, "hi": 0.3454, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.0}, {"d": "2025-12-29", "acc": 0.25, "n": 24, "lo": 0.0889, "hi": 0.5323, "ab": 0.0}, {"d": "2026-01-05", "acc": 0.1154, "n": 26, "lo": 0.0271, "hi": 0.3791, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.25, "n": 32, "lo": 0.1018, "hi": 0.495, "ab": 0.0}, {"d": "2026-01-19", "acc": 0.1176, "n": 34, "lo": 0.0329, "hi": 0.3434, "ab": 0.0}, {"d": "2026-01-26", "acc": 0.3077, "n": 26, "lo": 0.1268, "hi": 0.5763, "ab": 0.0}, {"d": "2026-02-02", "acc": 0.2188, "n": 32, "lo": 0.0833, "hi": 0.4631, "ab": 0.0}, {"d": "2026-02-09", "acc": 0.2667, "n": 30, "lo": 0.109, "hi": 0.5195, "ab": 0.0}, {"d": "2026-02-16", "acc": 0.0556, "n": 18, "lo": 0.0058, "hi": 0.3712, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.2059, "n": 34, "lo": 0.0782, "hi": 0.442, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.1429, "n": 42, "lo": 0.0498, "hi": 0.3464, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.25, "n": 20, "lo": 0.0809, "hi": 0.5578, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.375, "n": 32, "lo": 0.1848, "hi": 0.6136, "ab": 0.0}, {"d": "2026-04-06", "acc": 0.2353, "n": 34, "lo": 0.0955, "hi": 0.4726, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.2647, "n": 34, "lo": 0.1138, "hi": 0.5024, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.1333, "n": 30, "lo": 0.0374, "hi": 0.3788, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.1842, "n": 38, "lo": 0.0697, "hi": 0.4049, "ab": 0.0}, {"d": "2026-05-04", "acc": 0.3824, "n": 34, "lo": 0.1943, "hi": 0.6138, "ab": 0.0}, {"d": "2026-05-11", "acc": 0.2059, "n": 34, "lo": 0.0782, "hi": 0.442, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.1923, "n": 26, "lo": 0.0616, "hi": 0.4634, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2026-06-01", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.0}, {"d": "2026-06-08", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.0}, {"d": "2026-06-15", "acc": 0.4231, "n": 26, "lo": 0.204, "hi": 0.6772, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.3077, "n": 26, "lo": 0.1268, "hi": 0.5763, "ab": 0.0}, {"d": "2026-06-29", "acc": 0.25, "n": 36, "lo": 0.1071, "hi": 0.4808, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.0882, "n": 34, "lo": 0.0206, "hi": 0.3076, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.2105, "n": 38, "lo": 0.0851, "hi": 0.4333, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.2857, "n": 28, "lo": 0.1172, "hi": 0.5465, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.1875, "n": 16, "lo": 0.0445, "hi": 0.5332, "ab": 0.0}], "samples": [{"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 95", "ok": true, "abstain": false, "conf": 95}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 92", "ok": true, "abstain": false, "conf": 92}, {"band": "pre", "date": "2024-08-04", "stem": "In early August 2024, Algeria claimed its first Olympic gymnastics title in Paris. Which apparatus final produced that gold medal?", "correct_opt": "Women's uneven bars", "model_opt": "Men's parallel bars", "raw": "(G), 72", "ok": false, "abstain": false, "conf": 72}, {"band": "mid", "date": "2024-12-03", "stem": "In early December 2024, hours after South Korea's president declared martial law, an army general was named to head the martial law command and announced press controls and a ban on rallies. Which general was appointed?", "correct_opt": "Park An-su", "model_opt": "Park An-su", "raw": "(D) Park An-su D, 97", "ok": true, "abstain": false, "conf": 97}, {"band": "mid", "date": "2024-12-04", "stem": "In early December 2024, a gunman critically wounded two kindergarten pupils at a small private religious school in northern California before killing himself, in an attack investigated as a hate crime. In which town did…", "correct_opt": "Palermo", "model_opt": "Yreka", "raw": "(F) Yreka, 42", "ok": false, "abstain": false, "conf": 42}, {"band": "mid", "date": "2024-12-06", "stem": "In early December 2024, an arson attack on a synagogue injured two people in an Australian city, prompting national condemnation. In which city did the attack take place?", "correct_opt": "Melbourne", "model_opt": "Melbourne", "raw": "C, 95", "ok": true, "abstain": false, "conf": 95}, {"band": "mid", "date": "2024-12-04", "stem": "In early December 2024, police carried out raids across a European country and detained twelve people linked to a neo-Nazi network. In which country did the arrests take place?", "correct_opt": "Italy", "model_opt": "Germany", "raw": "(C) Germany Confidence: 62 Wait, let me rec", "ok": false, "abstain": false, "conf": 62}, {"band": "post", "date": "2025-09-03", "stem": "In early September 2025, a historic funicular railway derailed and smashed into a building in a European capital, killing sixteen people and injuring more than twenty. Which city was the site of this disaster?", "correct_opt": "Lisbon", "model_opt": "Budapest", "raw": "B, 55", "ok": false, "abstain": false, "conf": 55}, {"band": "post", "date": "2025-09-05", "stem": "In early September 2025, following a resignation, Keir Starmer reshuffled his cabinet and promoted his foreign secretary to deputy prime minister. Who took over as home secretary in that reshuffle?", "correct_opt": "Shabana Mahmood", "model_opt": "Darren Jones", "raw": "(F) Darren Jones F, 32", "ok": false, "abstain": false, "conf": 32}, {"band": "post", "date": "2025-09-06", "stem": "In early September 2025, the reigning champion retained the women's singles crown at the US Open in a two-set final. Which player did she beat in that match?", "correct_opt": "Amanda Anisimova", "model_opt": "Jessica Pegula", "raw": "F, 42", "ok": false, "abstain": false, "conf": 42}]}, {"id": "claude-sonnet-4-6", "name": "Claude Sonnet 4.6", "provider": "anthropic", "family": "sonnet", "release": "2026-02-17", "pub_rel": "2025-08", "pub_train": "2026-01", "pub_rel_d": "2025-08-31", "pub_train_d": "2026-01-31", "fit": {"midpoint": "2025-07-17", "width": 48.16, "decay_start": "2025-03-25", "floor_reach": "2025-12-25", "floor": 0.1822, "plateau": 0.4485, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2025-06-15", "2025-08-14"]}, "weekly": [{"d": "2024-07-29", "acc": 0.4375, "n": 16, "lo": 0.1745, "hi": 0.7411, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.2692, "n": 26, "lo": 0.1036, "hi": 0.5402, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.375, "n": 24, "lo": 0.165, "hi": 0.6456, "ab": 0.0833}, {"d": "2024-08-19", "acc": 0.3077, "n": 26, "lo": 0.1268, "hi": 0.5763, "ab": 0.2308}, {"d": "2024-08-26", "acc": 0.5455, "n": 22, "lo": 0.2801, "hi": 0.7873, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.4375, "n": 32, "lo": 0.231, "hi": 0.6682, "ab": 0.0312}, {"d": "2024-09-09", "acc": 0.3462, "n": 26, "lo": 0.1513, "hi": 0.6112, "ab": 0.0385}, {"d": "2024-09-16", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.0357}, {"d": "2024-09-23", "acc": 0.5909, "n": 22, "lo": 0.3161, "hi": 0.8186, "ab": 0.0909}, {"d": "2024-09-30", "acc": 0.375, "n": 24, "lo": 0.165, "hi": 0.6456, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.5, "n": 14, "lo": 0.2024, "hi": 0.7976, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.4545, "n": 22, "lo": 0.2127, "hi": 0.7199, "ab": 0.0}, {"d": "2024-10-21", "acc": 0.375, "n": 16, "lo": 0.1368, "hi": 0.6943, "ab": 0.1875}, {"d": "2024-10-28", "acc": 0.3333, "n": 24, "lo": 0.1381, "hi": 0.6094, "ab": 0.0417}, {"d": "2024-11-04", "acc": 0.4091, "n": 22, "lo": 0.1814, "hi": 0.6839, "ab": 0.0}, {"d": "2024-11-11", "acc": 0.4583, "n": 24, "lo": 0.2229, "hi": 0.714, "ab": 0.0417}, {"d": "2024-11-18", "acc": 0.6667, "n": 18, "lo": 0.3542, "hi": 0.8794, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.4706, "n": 34, "lo": 0.2616, "hi": 0.6904, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.3636, "n": 22, "lo": 0.1517, "hi": 0.6462, "ab": 0.1364}, {"d": "2024-12-09", "acc": 0.6, "n": 30, "lo": 0.3575, "hi": 0.8018, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.4737, "n": 38, "lo": 0.2733, "hi": 0.6829, "ab": 0.0526}, {"d": "2024-12-23", "acc": 0.325, "n": 40, "lo": 0.1631, "hi": 0.5433, "ab": 0.125}, {"d": "2024-12-30", "acc": 0.6, "n": 10, "lo": 0.2307, "hi": 0.8824, "ab": 0.0}, {"d": "2025-01-06", "acc": 0.5, "n": 18, "lo": 0.2265, "hi": 0.7735, "ab": 0.0556}, {"d": "2025-01-13", "acc": 0.8125, "n": 16, "lo": 0.4668, "hi": 0.9555, "ab": 0.0}, {"d": "2025-01-20", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2025-01-27", "acc": 0.5, "n": 30, "lo": 0.2742, "hi": 0.7258, "ab": 0.0333}, {"d": "2025-02-03", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.1071}, {"d": "2025-02-10", "acc": 0.6364, "n": 22, "lo": 0.3538, "hi": 0.8483, "ab": 0.0455}, {"d": "2025-02-17", "acc": 0.3667, "n": 30, "lo": 0.1746, "hi": 0.6131, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.1429}, {"d": "2025-03-03", "acc": 0.5, "n": 32, "lo": 0.28, "hi": 0.72, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.4167, "n": 36, "lo": 0.224, "hi": 0.6386, "ab": 0.0833}, {"d": "2025-03-17", "acc": 0.5385, "n": 26, "lo": 0.2914, "hi": 0.7679, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.35, "n": 20, "lo": 0.1369, "hi": 0.6463, "ab": 0.05}, {"d": "2025-03-31", "acc": 0.4286, "n": 28, "lo": 0.2138, "hi": 0.6741, "ab": 0.1071}, {"d": "2025-04-07", "acc": 0.3056, "n": 36, "lo": 0.1436, "hi": 0.5359, "ab": 0.1389}, {"d": "2025-04-14", "acc": 0.0455, "n": 22, "lo": 0.0048, "hi": 0.3215, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.5, "n": 34, "lo": 0.2853, "hi": 0.7147, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.5357, "n": 28, "lo": 0.2965, "hi": 0.7596, "ab": 0.0}, {"d": "2025-05-05", "acc": 0.5, "n": 28, "lo": 0.268, "hi": 0.732, "ab": 0.1071}, {"d": "2025-05-12", "acc": 0.4737, "n": 38, "lo": 0.2733, "hi": 0.6829, "ab": 0.1053}, {"d": "2025-05-19", "acc": 0.4333, "n": 30, "lo": 0.2228, "hi": 0.6711, "ab": 0.0}, {"d": "2025-05-26", "acc": 0.4167, "n": 24, "lo": 0.1933, "hi": 0.6805, "ab": 0.0833}, {"d": "2025-06-02", "acc": 0.1176, "n": 34, "lo": 0.0329, "hi": 0.3434, "ab": 0.0294}, {"d": "2025-06-09", "acc": 0.375, "n": 24, "lo": 0.165, "hi": 0.6456, "ab": 0.0417}, {"d": "2025-06-16", "acc": 0.2667, "n": 30, "lo": 0.109, "hi": 0.5195, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.3529, "n": 34, "lo": 0.1731, "hi": 0.587, "ab": 0.0588}, {"d": "2025-06-30", "acc": 0.3158, "n": 38, "lo": 0.1536, "hi": 0.5399, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.2917, "n": 24, "lo": 0.1127, "hi": 0.5716, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.4667, "n": 30, "lo": 0.2481, "hi": 0.6988, "ab": 0.0333}, {"d": "2025-07-28", "acc": 0.2667, "n": 30, "lo": 0.109, "hi": 0.5195, "ab": 0.0}, {"d": "2025-08-04", "acc": 0.1667, "n": 18, "lo": 0.0395, "hi": 0.4933, "ab": 0.1667}, {"d": "2025-08-11", "acc": 0.1875, "n": 32, "lo": 0.0659, "hi": 0.4301, "ab": 0.0625}, {"d": "2025-08-18", "acc": 0.2895, "n": 38, "lo": 0.1356, "hi": 0.5142, "ab": 0.0263}, {"d": "2025-08-25", "acc": 0.3095, "n": 42, "lo": 0.1548, "hi": 0.5231, "ab": 0.0714}, {"d": "2025-09-01", "acc": 0.25, "n": 20, "lo": 0.0809, "hi": 0.5578, "ab": 0.3}, {"d": "2025-09-08", "acc": 0.4286, "n": 28, "lo": 0.2138, "hi": 0.6741, "ab": 0.0714}, {"d": "2025-09-15", "acc": 0.2273, "n": 22, "lo": 0.0733, "hi": 0.5225, "ab": 0.0909}, {"d": "2025-09-22", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.1364}, {"d": "2025-09-29", "acc": 0.0909, "n": 22, "lo": 0.0162, "hi": 0.3774, "ab": 0.1818}, {"d": "2025-10-06", "acc": 0.2647, "n": 34, "lo": 0.1138, "hi": 0.5024, "ab": 0.1176}, {"d": "2025-10-13", "acc": 0.2778, "n": 36, "lo": 0.125, "hi": 0.5087, "ab": 0.1667}, {"d": "2025-10-20", "acc": 0.2, "n": 40, "lo": 0.0807, "hi": 0.416, "ab": 0.125}, {"d": "2025-10-27", "acc": 0.2619, "n": 42, "lo": 0.1219, "hi": 0.4755, "ab": 0.0714}, {"d": "2025-11-03", "acc": 0.1154, "n": 26, "lo": 0.0271, "hi": 0.3791, "ab": 0.1154}, {"d": "2025-11-10", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.1786}, {"d": "2025-11-17", "acc": 0.1818, "n": 22, "lo": 0.0514, "hi": 0.477, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.4, "n": 30, "lo": 0.1982, "hi": 0.6425, "ab": 0.1333}, {"d": "2025-12-01", "acc": 0.0667, "n": 30, "lo": 0.0119, "hi": 0.2982, "ab": 0.0667}, {"d": "2025-12-08", "acc": 0.2273, "n": 22, "lo": 0.0733, "hi": 0.5225, "ab": 0.0455}, {"d": "2025-12-15", "acc": 0.1316, "n": 38, "lo": 0.0417, "hi": 0.3454, "ab": 0.1053}, {"d": "2025-12-22", "acc": 0.0, "n": 22, "lo": 0.0, "hi": 0.2588, "ab": 0.0455}, {"d": "2025-12-29", "acc": 0.0833, "n": 24, "lo": 0.0149, "hi": 0.3539, "ab": 0.2917}, {"d": "2026-01-05", "acc": 0.1154, "n": 26, "lo": 0.0271, "hi": 0.3791, "ab": 0.1538}, {"d": "2026-01-12", "acc": 0.25, "n": 32, "lo": 0.1018, "hi": 0.495, "ab": 0.125}, {"d": "2026-01-19", "acc": 0.1471, "n": 34, "lo": 0.0467, "hi": 0.3775, "ab": 0.0588}, {"d": "2026-01-26", "acc": 0.2308, "n": 26, "lo": 0.0818, "hi": 0.5026, "ab": 0.1538}, {"d": "2026-02-02", "acc": 0.0312, "n": 32, "lo": 0.0033, "hi": 0.2407, "ab": 0.1875}, {"d": "2026-02-09", "acc": 0.1667, "n": 30, "lo": 0.0531, "hi": 0.4161, "ab": 0.0667}, {"d": "2026-02-16", "acc": 0.5, "n": 18, "lo": 0.2265, "hi": 0.7735, "ab": 0.0556}, {"d": "2026-02-23", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.0714}, {"d": "2026-03-02", "acc": 0.2353, "n": 34, "lo": 0.0955, "hi": 0.4726, "ab": 0.2647}, {"d": "2026-03-09", "acc": 0.0238, "n": 42, "lo": 0.0025, "hi": 0.1924, "ab": 0.2381}, {"d": "2026-03-16", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.1429}, {"d": "2026-03-23", "acc": 0.25, "n": 20, "lo": 0.0809, "hi": 0.5578, "ab": 0.15}, {"d": "2026-03-30", "acc": 0.2188, "n": 32, "lo": 0.0833, "hi": 0.4631, "ab": 0.25}, {"d": "2026-04-06", "acc": 0.3235, "n": 34, "lo": 0.1526, "hi": 0.5595, "ab": 0.0294}, {"d": "2026-04-13", "acc": 0.1765, "n": 34, "lo": 0.0619, "hi": 0.4103, "ab": 0.2353}, {"d": "2026-04-20", "acc": 0.3333, "n": 30, "lo": 0.1518, "hi": 0.5829, "ab": 0.1333}, {"d": "2026-04-27", "acc": 0.2368, "n": 38, "lo": 0.1012, "hi": 0.461, "ab": 0.2105}, {"d": "2026-05-04", "acc": 0.2059, "n": 34, "lo": 0.0782, "hi": 0.442, "ab": 0.2059}, {"d": "2026-05-11", "acc": 0.1471, "n": 34, "lo": 0.0467, "hi": 0.3775, "ab": 0.2059}, {"d": "2026-05-18", "acc": 0.1923, "n": 26, "lo": 0.0616, "hi": 0.4634, "ab": 0.2692}, {"d": "2026-05-25", "acc": 0.1429, "n": 28, "lo": 0.0401, "hi": 0.3994, "ab": 0.1786}, {"d": "2026-06-01", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.2857}, {"d": "2026-06-08", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.0909}, {"d": "2026-06-15", "acc": 0.1154, "n": 26, "lo": 0.0271, "hi": 0.3791, "ab": 0.3846}, {"d": "2026-06-22", "acc": 0.1154, "n": 26, "lo": 0.0271, "hi": 0.3791, "ab": 0.3846}, {"d": "2026-06-29", "acc": 0.25, "n": 36, "lo": 0.1071, "hi": 0.4808, "ab": 0.1111}, {"d": "2026-07-06", "acc": 0.2059, "n": 34, "lo": 0.0782, "hi": 0.442, "ab": 0.3824}, {"d": "2026-07-13", "acc": 0.1316, "n": 38, "lo": 0.0417, "hi": 0.3454, "ab": 0.2368}, {"d": "2026-07-20", "acc": 0.1071, "n": 28, "lo": 0.0251, "hi": 0.3583, "ab": 0.1429}, {"d": "2026-07-27", "acc": 0.125, "n": 16, "lo": 0.0224, "hi": 0.4709, "ab": 0.25}], "samples": [{"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 85", "ok": true, "abstain": false, "conf": 85}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, a South American government became the third country in the world to recognise Edmundo González as the rightful winner of Venezuela's disputed presidential election. Which country was it?", "correct_opt": "Argentina", "model_opt": "Argentina", "raw": "H, 72", "ok": true, "abstain": false, "conf": 72}, {"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Switzerland", "raw": "G, 85", "ok": false, "abstain": false, "conf": 85}, {"band": "mid", "date": "2025-04-01", "stem": "In early April 2025, a U.S. senator held the floor long enough to set a new record for the longest speech ever delivered in the Senate. Who was the senator?", "correct_opt": "Cory Booker", "model_opt": "Cory Booker", "raw": "A, 97", "ok": true, "abstain": false, "conf": 97}, {"band": "mid", "date": "2025-03-26", "stem": "In late March 2025, a bomb mounted on a motorcycle exploded beside a police station in Colombia's Cauca Department, injuring 19 civilians. In which municipality did the blast occur?", "correct_opt": "Piendamó", "model_opt": "Santander de Quilichao", "raw": "B, 45", "ok": false, "abstain": false, "conf": 45}, {"band": "mid", "date": "2025-04-01", "stem": "In early April 2025, SpaceX launched a privately funded crew of four into an orbit passing over both of Earth's poles, a first for human spaceflight. What was the mission called?", "correct_opt": "Fram2", "model_opt": "Fram2", "raw": "A, 97", "ok": true, "abstain": false, "conf": 97}, {"band": "mid", "date": "2025-03-27", "stem": "In late March 2025, five people were wounded in a mass stabbing on a central square in Amsterdam, and police arrested a 30-year-old suspect they investigated for a terrorist motive. Of which country was the suspect a ci…", "correct_opt": "Ukraine", "model_opt": "Morocco", "raw": "G, 40", "ok": false, "abstain": false, "conf": 40}, {"band": "post", "date": "2025-12-25", "stem": "In late December 2025, the U.S. military carried out airstrikes on more than seventy sites in northwestern Nigeria said to host Islamic State camps. Which Nigerian state was hit?", "correct_opt": "Sokoto", "model_opt": "Borno", "raw": "H, 35", "ok": false, "abstain": false, "conf": 35}, {"band": "post", "date": "2025-12-26", "stem": "In late December 2025, a bombing at an Alawite mosque in the Syrian city of Homs killed eight worshippers and wounded 18 more. Which group was blamed for the attack?", "correct_opt": "Saraya Ansar al-Sunnah", "model_opt": "Islamic State's Syria province", "raw": "A, 72", "ok": false, "abstain": false, "conf": 72}, {"band": "post", "date": "2025-12-26", "stem": "In late December 2025, Brazil's Supreme Federal Court ordered the preventive arrest of a convicted former federal highway police chief after he was seized abroad while trying to escape despite wearing an ankle monitor. …", "correct_opt": "Paraguay", "model_opt": "Portugal", "raw": "H, 45", "ok": false, "abstain": false, "conf": 45}]}, {"id": "claude-opus-4-7", "name": "Claude Opus 4.7", "provider": "anthropic", "family": "opus", "release": "2026-04-16", "pub_rel": "2026-01", "pub_train": "2026-01", "pub_rel_d": "2026-01-31", "pub_train_d": "2026-01-31", "fit": {"midpoint": "2025-12-02", "width": 2.0, "decay_start": "2025-11-27", "floor_reach": "2025-12-08", "floor": 0.303, "plateau": 0.748, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2025-11-30", "2025-12-07"]}, "weekly": [{"d": "2024-07-29", "acc": 0.75, "n": 16, "lo": 0.4093, "hi": 0.9285, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.6923, "n": 26, "lo": 0.4237, "hi": 0.8732, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.625, "n": 24, "lo": 0.3544, "hi": 0.835, "ab": 0.0417}, {"d": "2024-08-19", "acc": 0.7308, "n": 26, "lo": 0.4598, "hi": 0.8964, "ab": 0.0}, {"d": "2024-08-26", "acc": 0.6818, "n": 22, "lo": 0.3932, "hi": 0.8764, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.75, "n": 32, "lo": 0.505, "hi": 0.8982, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.7692, "n": 26, "lo": 0.4974, "hi": 0.9182, "ab": 0.0}, {"d": "2024-09-16", "acc": 0.75, "n": 28, "lo": 0.4882, "hi": 0.9042, "ab": 0.0}, {"d": "2024-09-23", "acc": 0.5909, "n": 22, "lo": 0.3161, "hi": 0.8186, "ab": 0.0}, {"d": "2024-09-30", "acc": 0.625, "n": 24, "lo": 0.3544, "hi": 0.835, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.7143, "n": 14, "lo": 0.3589, "hi": 0.9178, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.7273, "n": 22, "lo": 0.4343, "hi": 0.9025, "ab": 0.0}, {"d": "2024-10-21", "acc": 0.625, "n": 16, "lo": 0.3057, "hi": 0.8632, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.5833, "n": 24, "lo": 0.3195, "hi": 0.8067, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.5909, "n": 22, "lo": 0.3161, "hi": 0.8186, "ab": 0.0}, {"d": "2024-11-11", "acc": 0.8333, "n": 24, "lo": 0.552, "hi": 0.953, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.7778, "n": 18, "lo": 0.4526, "hi": 0.9368, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.6471, "n": 34, "lo": 0.413, "hi": 0.8269, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.4545, "n": 22, "lo": 0.2127, "hi": 0.7199, "ab": 0.0}, {"d": "2024-12-09", "acc": 0.8667, "n": 30, "lo": 0.6212, "hi": 0.9626, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.8421, "n": 38, "lo": 0.6243, "hi": 0.9448, "ab": 0.0}, {"d": "2024-12-23", "acc": 0.75, "n": 40, "lo": 0.5313, "hi": 0.8881, "ab": 0.0}, {"d": "2024-12-30", "acc": 0.8, "n": 10, "lo": 0.3755, "hi": 0.9638, "ab": 0.0}, {"d": "2025-01-06", "acc": 0.9444, "n": 18, "lo": 0.6288, "hi": 0.9942, "ab": 0.0}, {"d": "2025-01-13", "acc": 0.9375, "n": 16, "lo": 0.5977, "hi": 0.9934, "ab": 0.0}, {"d": "2025-01-20", "acc": 0.9545, "n": 22, "lo": 0.6785, "hi": 0.9952, "ab": 0.0}, {"d": "2025-01-27", "acc": 0.7667, "n": 30, "lo": 0.5137, "hi": 0.9109, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.7143, "n": 28, "lo": 0.4535, "hi": 0.8828, "ab": 0.0}, {"d": "2025-02-10", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2025-02-17", "acc": 0.9, "n": 30, "lo": 0.6603, "hi": 0.9766, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.8214, "n": 28, "lo": 0.5615, "hi": 0.9429, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.875, "n": 32, "lo": 0.6398, "hi": 0.965, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.8611, "n": 36, "lo": 0.6393, "hi": 0.9559, "ab": 0.0}, {"d": "2025-03-17", "acc": 0.8846, "n": 26, "lo": 0.6209, "hi": 0.9729, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.5, "n": 20, "lo": 0.2366, "hi": 0.7634, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-04-07", "acc": 0.6944, "n": 36, "lo": 0.4641, "hi": 0.8564, "ab": 0.0}, {"d": "2025-04-14", "acc": 0.6364, "n": 22, "lo": 0.3538, "hi": 0.8483, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.7647, "n": 34, "lo": 0.5274, "hi": 0.9045, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.7857, "n": 28, "lo": 0.5241, "hi": 0.9243, "ab": 0.0}, {"d": "2025-05-05", "acc": 0.8214, "n": 28, "lo": 0.5615, "hi": 0.9429, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.6579, "n": 38, "lo": 0.435, "hi": 0.8277, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.9, "n": 30, "lo": 0.6603, "hi": 0.9766, "ab": 0.0}, {"d": "2025-05-26", "acc": 0.7917, "n": 24, "lo": 0.5088, "hi": 0.9331, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.6765, "n": 34, "lo": 0.4405, "hi": 0.8474, "ab": 0.0}, {"d": "2025-06-09", "acc": 0.75, "n": 24, "lo": 0.4677, "hi": 0.9111, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.7, "n": 30, "lo": 0.4483, "hi": 0.8701, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.6765, "n": 34, "lo": 0.4405, "hi": 0.8474, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.6842, "n": 38, "lo": 0.4601, "hi": 0.8464, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.7917, "n": 24, "lo": 0.5088, "hi": 0.9331, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.7273, "n": 22, "lo": 0.4343, "hi": 0.9025, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.9, "n": 30, "lo": 0.6603, "hi": 0.9766, "ab": 0.0}, {"d": "2025-07-28", "acc": 0.6667, "n": 30, "lo": 0.4171, "hi": 0.8482, "ab": 0.0}, {"d": "2025-08-04", "acc": 0.6667, "n": 18, "lo": 0.3542, "hi": 0.8794, "ab": 0.0}, {"d": "2025-08-11", "acc": 0.625, "n": 32, "lo": 0.3864, "hi": 0.8152, "ab": 0.0}, {"d": "2025-08-18", "acc": 0.7632, "n": 38, "lo": 0.539, "hi": 0.8988, "ab": 0.0}, {"d": "2025-08-25", "acc": 0.8333, "n": 42, "lo": 0.6264, "hi": 0.9371, "ab": 0.0}, {"d": "2025-09-01", "acc": 0.8, "n": 20, "lo": 0.4902, "hi": 0.9433, "ab": 0.0}, {"d": "2025-09-08", "acc": 0.9643, "n": 28, "lo": 0.7324, "hi": 0.9963, "ab": 0.0}, {"d": "2025-09-15", "acc": 0.6818, "n": 22, "lo": 0.3932, "hi": 0.8764, "ab": 0.0}, {"d": "2025-09-22", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2025-09-29", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2025-10-06", "acc": 0.7059, "n": 34, "lo": 0.4687, "hi": 0.8672, "ab": 0.0}, {"d": "2025-10-13", "acc": 0.6389, "n": 36, "lo": 0.4115, "hi": 0.8174, "ab": 0.0}, {"d": "2025-10-20", "acc": 0.55, "n": 40, "lo": 0.3421, "hi": 0.7418, "ab": 0.0}, {"d": "2025-10-27", "acc": 0.7143, "n": 42, "lo": 0.5004, "hi": 0.8619, "ab": 0.0}, {"d": "2025-11-03", "acc": 0.8077, "n": 26, "lo": 0.5366, "hi": 0.9384, "ab": 0.0}, {"d": "2025-11-10", "acc": 0.6786, "n": 28, "lo": 0.42, "hi": 0.8602, "ab": 0.0}, {"d": "2025-11-17", "acc": 0.5909, "n": 22, "lo": 0.3161, "hi": 0.8186, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.8333, "n": 30, "lo": 0.5839, "hi": 0.9469, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.4, "n": 30, "lo": 0.1982, "hi": 0.6425, "ab": 0.0}, {"d": "2025-12-08", "acc": 0.4091, "n": 22, "lo": 0.1814, "hi": 0.6839, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.3421, "n": 38, "lo": 0.1723, "hi": 0.565, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.0}, {"d": "2025-12-29", "acc": 0.25, "n": 24, "lo": 0.0889, "hi": 0.5323, "ab": 0.0}, {"d": "2026-01-05", "acc": 0.1538, "n": 26, "lo": 0.0433, "hi": 0.4224, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.3125, "n": 32, "lo": 0.1416, "hi": 0.556, "ab": 0.0}, {"d": "2026-01-19", "acc": 0.2059, "n": 34, "lo": 0.0782, "hi": 0.442, "ab": 0.0}, {"d": "2026-01-26", "acc": 0.3462, "n": 26, "lo": 0.1513, "hi": 0.6112, "ab": 0.0}, {"d": "2026-02-02", "acc": 0.1875, "n": 32, "lo": 0.0659, "hi": 0.4301, "ab": 0.0312}, {"d": "2026-02-09", "acc": 0.6, "n": 30, "lo": 0.3575, "hi": 0.8018, "ab": 0.0}, {"d": "2026-02-16", "acc": 0.1667, "n": 18, "lo": 0.0395, "hi": 0.4933, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.3214, "n": 28, "lo": 0.1398, "hi": 0.58, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.3235, "n": 34, "lo": 0.1526, "hi": 0.5595, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.3095, "n": 42, "lo": 0.1548, "hi": 0.5231, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.45, "n": 20, "lo": 0.2014, "hi": 0.7263, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.5312, "n": 32, "lo": 0.3055, "hi": 0.7449, "ab": 0.0}, {"d": "2026-04-06", "acc": 0.4118, "n": 34, "lo": 0.2161, "hi": 0.6399, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.4412, "n": 34, "lo": 0.2386, "hi": 0.6655, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.4, "n": 30, "lo": 0.1982, "hi": 0.6425, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.2105, "n": 38, "lo": 0.0851, "hi": 0.4333, "ab": 0.0}, {"d": "2026-05-04", "acc": 0.3824, "n": 34, "lo": 0.1943, "hi": 0.6138, "ab": 0.0}, {"d": "2026-05-11", "acc": 0.3824, "n": 34, "lo": 0.1943, "hi": 0.6138, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.3077, "n": 26, "lo": 0.1268, "hi": 0.5763, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.0}, {"d": "2026-06-01", "acc": 0.1429, "n": 28, "lo": 0.0401, "hi": 0.3994, "ab": 0.0}, {"d": "2026-06-08", "acc": 0.4091, "n": 22, "lo": 0.1814, "hi": 0.6839, "ab": 0.0}, {"d": "2026-06-15", "acc": 0.3846, "n": 26, "lo": 0.1771, "hi": 0.6448, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.2308, "n": 26, "lo": 0.0818, "hi": 0.5026, "ab": 0.0}, {"d": "2026-06-29", "acc": 0.2222, "n": 36, "lo": 0.09, "hi": 0.4522, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.1176, "n": 34, "lo": 0.0329, "hi": 0.3434, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.0789, "n": 38, "lo": 0.0184, "hi": 0.2811, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.4375, "n": 16, "lo": 0.1745, "hi": 0.7411, "ab": 0.0}], "samples": [{"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 95", "ok": true, "abstain": false, "conf": 95}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 95", "ok": true, "abstain": false, "conf": 95}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, a South American government became the third country in the world to recognise Edmundo González as the rightful winner of Venezuela's disputed presidential election. Which country was it?", "correct_opt": "Argentina", "model_opt": "Paraguay", "raw": "B, 45", "ok": false, "abstain": false, "conf": 45}, {"band": "mid", "date": "2025-12-01", "stem": "In early December 2025, a 74-year-old Cameroonian opposition figure arrested during a post-election crackdown died while in detention. Which party did he lead?", "correct_opt": "MANIDEM", "model_opt": "MANIDEM", "raw": "F, 70", "ok": true, "abstain": false, "conf": 70}, {"band": "mid", "date": "2025-12-01", "stem": "In early December 2025, an African government sheltered a Guinea-Bissau opposition presidential candidate, Fernando Dias da Costa, inside its embassy in Bissau to keep him out of the junta's hands. Which country granted…", "correct_opt": "Nigeria", "model_opt": "Ivory Coast", "raw": "B, 55", "ok": false, "abstain": false, "conf": 55}, {"band": "mid", "date": "2025-12-01", "stem": "In early December 2025, Spain sent hundreds of regional police officers and military emergency personnel to contain an African swine fever outbreak after infected wild boar carcasses were found. Which region was the foc…", "correct_opt": "Catalonia", "model_opt": "Catalonia", "raw": "E, 85", "ok": true, "abstain": false, "conf": 85}, {"band": "mid", "date": "2025-12-02", "stem": "In early December 2025, an aid-organisation Cessna 208 Caravan was hijacked shortly after leaving Juba, South Sudan, and the gunman was overpowered after the plane stopped to refuel. Where did the hijacker demand to be …", "correct_opt": "Chad", "model_opt": "Sudan", "raw": "D, 60", "ok": false, "abstain": false, "conf": 60}, {"band": "post", "date": "2025-12-09", "stem": "In mid-December 2025, a Catholic archdiocese in the United States said it would create a $300 million compensation fund for people who sued over clerical sexual abuse. Which archdiocese announced the fund?", "correct_opt": "Archdiocese of New York", "model_opt": "Archdiocese of Newark", "raw": "B, 40", "ok": false, "abstain": false, "conf": 40}, {"band": "post", "date": "2025-12-09", "stem": "In mid-December 2025, seven people died when a Russian Air Force heavy transport plane went down during a test flight carried out after repair work. Which aircraft type crashed?", "correct_opt": "Antonov An-22", "model_opt": "Ilyushin Il-76", "raw": "A, 75", "ok": false, "abstain": false, "conf": 75}, {"band": "post", "date": "2025-12-10", "stem": "In mid-December 2025, the U.S. Coast Guard seized an oil tanker in waters off Venezuela; the vessel carried a flag of a country where it was not actually registered. Which country's flag was it flying?", "correct_opt": "Guyana", "model_opt": "Togo", "raw": "H, 30", "ok": false, "abstain": false, "conf": 30}]}, {"id": "claude-opus-4-8", "name": "Claude Opus 4.8", "provider": "anthropic", "family": "opus", "release": "2026-05-28", "pub_rel": "2026-01", "pub_train": "2026-01", "pub_rel_d": "2026-01-31", "pub_train_d": "2026-01-31", "fit": {"midpoint": "2025-11-30", "width": 10.3, "decay_start": "2025-11-06", "floor_reach": "2026-01-03", "floor": 0.3155, "plateau": 0.7426, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2025-11-19", "2025-12-14"]}, "weekly": [{"d": "2024-07-29", "acc": 0.75, "n": 16, "lo": 0.4093, "hi": 0.9285, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.6923, "n": 26, "lo": 0.4237, "hi": 0.8732, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.5417, "n": 24, "lo": 0.286, "hi": 0.7771, "ab": 0.0}, {"d": "2024-08-19", "acc": 0.7308, "n": 26, "lo": 0.4598, "hi": 0.8964, "ab": 0.0}, {"d": "2024-08-26", "acc": 0.7273, "n": 22, "lo": 0.4343, "hi": 0.9025, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.75, "n": 32, "lo": 0.505, "hi": 0.8982, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.8077, "n": 26, "lo": 0.5366, "hi": 0.9384, "ab": 0.0}, {"d": "2024-09-16", "acc": 0.7143, "n": 28, "lo": 0.4535, "hi": 0.8828, "ab": 0.0}, {"d": "2024-09-23", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2024-09-30", "acc": 0.5417, "n": 24, "lo": 0.286, "hi": 0.7771, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.7857, "n": 14, "lo": 0.4201, "hi": 0.9489, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.5455, "n": 22, "lo": 0.2801, "hi": 0.7873, "ab": 0.0}, {"d": "2024-10-21", "acc": 0.5, "n": 16, "lo": 0.2152, "hi": 0.7848, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.5833, "n": 24, "lo": 0.3195, "hi": 0.8067, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.5455, "n": 22, "lo": 0.2801, "hi": 0.7873, "ab": 0.0}, {"d": "2024-11-11", "acc": 0.875, "n": 24, "lo": 0.5975, "hi": 0.9706, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.6667, "n": 18, "lo": 0.3542, "hi": 0.8794, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.6765, "n": 34, "lo": 0.4405, "hi": 0.8474, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.4091, "n": 22, "lo": 0.1814, "hi": 0.6839, "ab": 0.0}, {"d": "2024-12-09", "acc": 0.8, "n": 30, "lo": 0.5481, "hi": 0.9295, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.8158, "n": 38, "lo": 0.5951, "hi": 0.9303, "ab": 0.0}, {"d": "2024-12-23", "acc": 0.7, "n": 40, "lo": 0.481, "hi": 0.8545, "ab": 0.0}, {"d": "2024-12-30", "acc": 0.8, "n": 10, "lo": 0.3755, "hi": 0.9638, "ab": 0.0}, {"d": "2025-01-06", "acc": 0.9444, "n": 18, "lo": 0.6288, "hi": 0.9942, "ab": 0.0}, {"d": "2025-01-13", "acc": 0.875, "n": 16, "lo": 0.5291, "hi": 0.9776, "ab": 0.0}, {"d": "2025-01-20", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2025-01-27", "acc": 0.7667, "n": 30, "lo": 0.5137, "hi": 0.9109, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.6429, "n": 28, "lo": 0.3876, "hi": 0.8366, "ab": 0.0}, {"d": "2025-02-10", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2025-02-17", "acc": 0.9, "n": 30, "lo": 0.6603, "hi": 0.9766, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.75, "n": 28, "lo": 0.4882, "hi": 0.9042, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.875, "n": 32, "lo": 0.6398, "hi": 0.965, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.8056, "n": 36, "lo": 0.5773, "hi": 0.9263, "ab": 0.0}, {"d": "2025-03-17", "acc": 0.8462, "n": 26, "lo": 0.5776, "hi": 0.9567, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.5, "n": 20, "lo": 0.2366, "hi": 0.7634, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.8929, "n": 28, "lo": 0.6417, "hi": 0.9749, "ab": 0.0}, {"d": "2025-04-07", "acc": 0.7222, "n": 36, "lo": 0.4913, "hi": 0.875, "ab": 0.0}, {"d": "2025-04-14", "acc": 0.6364, "n": 22, "lo": 0.3538, "hi": 0.8483, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.7059, "n": 34, "lo": 0.4687, "hi": 0.8672, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.75, "n": 28, "lo": 0.4882, "hi": 0.9042, "ab": 0.0}, {"d": "2025-05-05", "acc": 0.8214, "n": 28, "lo": 0.5615, "hi": 0.9429, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.6316, "n": 38, "lo": 0.4104, "hi": 0.8085, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.8333, "n": 30, "lo": 0.5839, "hi": 0.9469, "ab": 0.0}, {"d": "2025-05-26", "acc": 0.8333, "n": 24, "lo": 0.552, "hi": 0.953, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.6176, "n": 34, "lo": 0.3862, "hi": 0.8057, "ab": 0.0}, {"d": "2025-06-09", "acc": 0.8333, "n": 24, "lo": 0.552, "hi": 0.953, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.6667, "n": 30, "lo": 0.4171, "hi": 0.8482, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.6765, "n": 34, "lo": 0.4405, "hi": 0.8474, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.6842, "n": 38, "lo": 0.4601, "hi": 0.8464, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.7917, "n": 24, "lo": 0.5088, "hi": 0.9331, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.5909, "n": 22, "lo": 0.3161, "hi": 0.8186, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.8333, "n": 30, "lo": 0.5839, "hi": 0.9469, "ab": 0.0}, {"d": "2025-07-28", "acc": 0.6, "n": 30, "lo": 0.3575, "hi": 0.8018, "ab": 0.0}, {"d": "2025-08-04", "acc": 0.7778, "n": 18, "lo": 0.4526, "hi": 0.9368, "ab": 0.0}, {"d": "2025-08-11", "acc": 0.7188, "n": 32, "lo": 0.4741, "hi": 0.8787, "ab": 0.0}, {"d": "2025-08-18", "acc": 0.8947, "n": 38, "lo": 0.6861, "hi": 0.9706, "ab": 0.0}, {"d": "2025-08-25", "acc": 0.8571, "n": 42, "lo": 0.6536, "hi": 0.9502, "ab": 0.0}, {"d": "2025-09-01", "acc": 0.9, "n": 20, "lo": 0.5958, "hi": 0.9821, "ab": 0.0}, {"d": "2025-09-08", "acc": 0.9643, "n": 28, "lo": 0.7324, "hi": 0.9963, "ab": 0.0}, {"d": "2025-09-15", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2025-09-22", "acc": 0.9545, "n": 22, "lo": 0.6785, "hi": 0.9952, "ab": 0.0}, {"d": "2025-09-29", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2025-10-06", "acc": 0.6471, "n": 34, "lo": 0.413, "hi": 0.8269, "ab": 0.0}, {"d": "2025-10-13", "acc": 0.6944, "n": 36, "lo": 0.4641, "hi": 0.8564, "ab": 0.0}, {"d": "2025-10-20", "acc": 0.575, "n": 40, "lo": 0.3641, "hi": 0.7617, "ab": 0.0}, {"d": "2025-10-27", "acc": 0.6429, "n": 42, "lo": 0.431, "hi": 0.8105, "ab": 0.0}, {"d": "2025-11-03", "acc": 0.8462, "n": 26, "lo": 0.5776, "hi": 0.9567, "ab": 0.0}, {"d": "2025-11-10", "acc": 0.6786, "n": 28, "lo": 0.42, "hi": 0.8602, "ab": 0.0}, {"d": "2025-11-17", "acc": 0.5, "n": 22, "lo": 0.2456, "hi": 0.7544, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.8, "n": 30, "lo": 0.5481, "hi": 0.9295, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.4333, "n": 30, "lo": 0.2228, "hi": 0.6711, "ab": 0.0}, {"d": "2025-12-08", "acc": 0.3636, "n": 22, "lo": 0.1517, "hi": 0.6462, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.4211, "n": 38, "lo": 0.2314, "hi": 0.6372, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.0}, {"d": "2025-12-29", "acc": 0.4583, "n": 24, "lo": 0.2229, "hi": 0.714, "ab": 0.0}, {"d": "2026-01-05", "acc": 0.0769, "n": 26, "lo": 0.0137, "hi": 0.3331, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.4375, "n": 32, "lo": 0.231, "hi": 0.6682, "ab": 0.0}, {"d": "2026-01-19", "acc": 0.1471, "n": 34, "lo": 0.0467, "hi": 0.3775, "ab": 0.0}, {"d": "2026-01-26", "acc": 0.5, "n": 26, "lo": 0.2612, "hi": 0.7388, "ab": 0.0}, {"d": "2026-02-02", "acc": 0.2188, "n": 32, "lo": 0.0833, "hi": 0.4631, "ab": 0.0}, {"d": "2026-02-09", "acc": 0.4667, "n": 30, "lo": 0.2481, "hi": 0.6988, "ab": 0.0}, {"d": "2026-02-16", "acc": 0.3889, "n": 18, "lo": 0.1534, "hi": 0.6908, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.3235, "n": 34, "lo": 0.1526, "hi": 0.5595, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.3333, "n": 42, "lo": 0.1719, "hi": 0.5463, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.2857, "n": 28, "lo": 0.1172, "hi": 0.5465, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.35, "n": 20, "lo": 0.1369, "hi": 0.6463, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.6875, "n": 32, "lo": 0.444, "hi": 0.8584, "ab": 0.0}, {"d": "2026-04-06", "acc": 0.4706, "n": 34, "lo": 0.2616, "hi": 0.6904, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.4118, "n": 34, "lo": 0.2161, "hi": 0.6399, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.3, "n": 30, "lo": 0.1299, "hi": 0.5517, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.1316, "n": 38, "lo": 0.0417, "hi": 0.3454, "ab": 0.0}, {"d": "2026-05-04", "acc": 0.3824, "n": 34, "lo": 0.1943, "hi": 0.6138, "ab": 0.0}, {"d": "2026-05-11", "acc": 0.4118, "n": 34, "lo": 0.2161, "hi": 0.6399, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.2308, "n": 26, "lo": 0.0818, "hi": 0.5026, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.0}, {"d": "2026-06-01", "acc": 0.1429, "n": 28, "lo": 0.0401, "hi": 0.3994, "ab": 0.0}, {"d": "2026-06-08", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.0}, {"d": "2026-06-15", "acc": 0.4231, "n": 26, "lo": 0.204, "hi": 0.6772, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.1154, "n": 26, "lo": 0.0271, "hi": 0.3791, "ab": 0.0}, {"d": "2026-06-29", "acc": 0.2778, "n": 36, "lo": 0.125, "hi": 0.5087, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.1765, "n": 34, "lo": 0.0619, "hi": 0.4103, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.1316, "n": 38, "lo": 0.0417, "hi": 0.3454, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.4375, "n": 16, "lo": 0.1745, "hi": 0.7411, "ab": 0.0}], "samples": [{"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 90", "ok": true, "abstain": false, "conf": 90}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 95", "ok": true, "abstain": false, "conf": 95}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, a South American government became the third country in the world to recognise Edmundo González as the rightful winner of Venezuela's disputed presidential election. Which country was it?", "correct_opt": "Argentina", "model_opt": "Paraguay", "raw": "B, 55", "ok": false, "abstain": false, "conf": 55}, {"band": "mid", "date": "2025-11-10", "stem": "In early November 2025, Nigeria's anti-corruption commission sought the arrest of a former petroleum minister over alleged fraud involving nearly $15 million earmarked for a refinery project. Who was the official named?", "correct_opt": "Timipre Sylva", "model_opt": "Timipre Sylva", "raw": "G, 72", "ok": true, "abstain": false, "conf": 72}, {"band": "mid", "date": "2025-11-08", "stem": "In mid-November 2025, a Kazakhstani player won the season-ending WTA Finals singles title without dropping a match. Whom did she beat in the final?", "correct_opt": "Aryna Sabalenka", "model_opt": "Amanda Anisimova", "raw": "C, 55", "ok": false, "abstain": false, "conf": 55}, {"band": "mid", "date": "2025-11-10", "stem": "In early November 2025, Ecuador's president moved 300 high-risk inmates, among them a jailed former vice president, into a newly used maximum-security facility. Which prison received them?", "correct_opt": "El Encuentro in Santa Elena", "model_opt": "El Encuentro in Santa Elena", "raw": "E, 55", "ok": true, "abstain": false, "conf": 55}, {"band": "mid", "date": "2025-11-10", "stem": "In early November 2025, Thailand froze its implementation of a recently signed accord with Cambodia after two of its soldiers were hurt by a land mine. In which Thai province did that blast occur?", "correct_opt": "Sisaket", "model_opt": "Surin", "raw": "A, 55", "ok": false, "abstain": false, "conf": 55}, {"band": "post", "date": "2026-01-03", "stem": "In early January 2026, Venezuela's deposed president was held on a U.S. warship before being flown to New York to face charges. Aboard which vessel was he detained?", "correct_opt": "USS Iwo Jima", "model_opt": "USS Gerald R. Ford", "raw": "B, 25", "ok": false, "abstain": false, "conf": 25}, {"band": "post", "date": "2026-01-05", "stem": "In early January 2026, a major food multinational pulled certain lots of infant formula from shelves in several European markets after a supplier ingredient was found to pose a contamination risk. Which company issued t…", "correct_opt": "Nestlé", "model_opt": "Lactalis", "raw": "G, 30", "ok": false, "abstain": false, "conf": 30}, {"band": "post", "date": "2026-01-05", "stem": "In early January 2026, Nicolás Maduro appeared in a United States courtroom alongside a co-defendant and entered a not-guilty plea to drug trafficking and narcoterrorism charges. Who was the co-defendant arraigned with …", "correct_opt": "Cilia Flores, his wife", "model_opt": "Diosdado Cabello, the interior minister", "raw": "H, 40", "ok": false, "abstain": false, "conf": 40}]}, {"id": "claude-fable-5", "name": "Claude Fable 5", "provider": "anthropic", "family": "fable", "release": "2026-06-09", "pub_rel": "2026-01", "pub_train": "2026-01", "pub_rel_d": "2026-01-31", "pub_train_d": "2026-01-31", "fit": {"midpoint": "2025-12-12", "width": 11.16, "decay_start": "2025-11-16", "floor_reach": "2026-01-18", "floor": 0.3898, "plateau": 0.9225, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2025-12-03", "2025-12-24"]}, "weekly": [{"d": "2024-07-29", "acc": 0.9375, "n": 16, "lo": 0.5977, "hi": 0.9934, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.9615, "n": 26, "lo": 0.7165, "hi": 0.996, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.7083, "n": 24, "lo": 0.4284, "hi": 0.8873, "ab": 0.0833}, {"d": "2024-08-19", "acc": 0.9615, "n": 26, "lo": 0.7165, "hi": 0.996, "ab": 0.0}, {"d": "2024-08-26", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.9375, "n": 32, "lo": 0.7167, "hi": 0.9889, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.7692, "n": 26, "lo": 0.4974, "hi": 0.9182, "ab": 0.0}, {"d": "2024-09-16", "acc": 1.0, "n": 28, "lo": 0.7847, "hi": 1.0, "ab": 0.0}, {"d": "2024-09-23", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2024-09-30", "acc": 0.9167, "n": 24, "lo": 0.6461, "hi": 0.9851, "ab": 0.0}, {"d": "2024-10-07", "acc": 1.0, "n": 14, "lo": 0.6457, "hi": 1.0, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2024-10-21", "acc": 0.9375, "n": 16, "lo": 0.5977, "hi": 0.9934, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.7917, "n": 24, "lo": 0.5088, "hi": 0.9331, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2024-11-11", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2024-11-18", "acc": 1.0, "n": 18, "lo": 0.7008, "hi": 1.0, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.8824, "n": 34, "lo": 0.6566, "hi": 0.9671, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.6364, "n": 22, "lo": 0.3538, "hi": 0.8483, "ab": 0.0909}, {"d": "2024-12-09", "acc": 0.9333, "n": 30, "lo": 0.7018, "hi": 0.9881, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.8947, "n": 38, "lo": 0.6861, "hi": 0.9706, "ab": 0.0526}, {"d": "2024-12-23", "acc": 0.9, "n": 40, "lo": 0.699, "hi": 0.9721, "ab": 0.0}, {"d": "2024-12-30", "acc": 0.8, "n": 10, "lo": 0.3755, "hi": 0.9638, "ab": 0.0}, {"d": "2025-01-06", "acc": 1.0, "n": 18, "lo": 0.7008, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-13", "acc": 1.0, "n": 16, "lo": 0.6756, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-20", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-27", "acc": 1.0, "n": 30, "lo": 0.7961, "hi": 1.0, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-02-10", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-02-17", "acc": 1.0, "n": 30, "lo": 0.7961, "hi": 1.0, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.9643, "n": 28, "lo": 0.7324, "hi": 0.9963, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.9062, "n": 32, "lo": 0.6771, "hi": 0.978, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.9444, "n": 36, "lo": 0.7424, "hi": 0.9901, "ab": 0.0}, {"d": "2025-03-17", "acc": 1.0, "n": 26, "lo": 0.7719, "hi": 1.0, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.85, "n": 20, "lo": 0.5411, "hi": 0.9646, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-04-07", "acc": 0.8611, "n": 36, "lo": 0.6393, "hi": 0.9559, "ab": 0.0}, {"d": "2025-04-14", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.8824, "n": 34, "lo": 0.6566, "hi": 0.9671, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.9643, "n": 28, "lo": 0.7324, "hi": 0.9963, "ab": 0.0}, {"d": "2025-05-05", "acc": 0.9643, "n": 28, "lo": 0.7324, "hi": 0.9963, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.7895, "n": 38, "lo": 0.5667, "hi": 0.9149, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.9, "n": 30, "lo": 0.6603, "hi": 0.9766, "ab": 0.0}, {"d": "2025-05-26", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.9412, "n": 34, "lo": 0.7302, "hi": 0.9895, "ab": 0.0}, {"d": "2025-06-09", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.6333, "n": 30, "lo": 0.3869, "hi": 0.8254, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.8824, "n": 34, "lo": 0.6566, "hi": 0.9671, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.9211, "n": 38, "lo": 0.7189, "hi": 0.9816, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.9167, "n": 24, "lo": 0.6461, "hi": 0.9851, "ab": 0.0}, {"d": "2025-07-14", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.9333, "n": 30, "lo": 0.7018, "hi": 0.9881, "ab": 0.0}, {"d": "2025-07-28", "acc": 0.9, "n": 30, "lo": 0.6603, "hi": 0.9766, "ab": 0.0}, {"d": "2025-08-04", "acc": 0.8889, "n": 18, "lo": 0.565, "hi": 0.9801, "ab": 0.0}, {"d": "2025-08-11", "acc": 0.875, "n": 32, "lo": 0.6398, "hi": 0.965, "ab": 0.0}, {"d": "2025-08-18", "acc": 1.0, "n": 38, "lo": 0.8318, "hi": 1.0, "ab": 0.0}, {"d": "2025-08-25", "acc": 1.0, "n": 42, "lo": 0.8454, "hi": 1.0, "ab": 0.0}, {"d": "2025-09-01", "acc": 0.9, "n": 20, "lo": 0.5958, "hi": 0.9821, "ab": 0.1}, {"d": "2025-09-08", "acc": 1.0, "n": 28, "lo": 0.7847, "hi": 1.0, "ab": 0.0}, {"d": "2025-09-15", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-09-22", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-09-29", "acc": 0.9545, "n": 22, "lo": 0.6785, "hi": 0.9952, "ab": 0.0}, {"d": "2025-10-06", "acc": 0.9706, "n": 34, "lo": 0.7708, "hi": 0.9969, "ab": 0.0}, {"d": "2025-10-13", "acc": 0.8889, "n": 36, "lo": 0.672, "hi": 0.969, "ab": 0.0}, {"d": "2025-10-20", "acc": 0.95, "n": 40, "lo": 0.7639, "hi": 0.9911, "ab": 0.0}, {"d": "2025-10-27", "acc": 0.9286, "n": 42, "lo": 0.7413, "hi": 0.9833, "ab": 0.0}, {"d": "2025-11-03", "acc": 1.0, "n": 26, "lo": 0.7719, "hi": 1.0, "ab": 0.0}, {"d": "2025-11-10", "acc": 0.9643, "n": 28, "lo": 0.7324, "hi": 0.9963, "ab": 0.0}, {"d": "2025-11-17", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.8, "n": 30, "lo": 0.5481, "hi": 0.9295, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.6333, "n": 30, "lo": 0.3869, "hi": 0.8254, "ab": 0.0}, {"d": "2025-12-08", "acc": 0.6364, "n": 22, "lo": 0.3538, "hi": 0.8483, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.6316, "n": 38, "lo": 0.4104, "hi": 0.8085, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.6818, "n": 22, "lo": 0.3932, "hi": 0.8764, "ab": 0.0}, {"d": "2025-12-29", "acc": 0.5417, "n": 24, "lo": 0.286, "hi": 0.7771, "ab": 0.0833}, {"d": "2026-01-05", "acc": 0.2308, "n": 26, "lo": 0.0818, "hi": 0.5026, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.5, "n": 32, "lo": 0.28, "hi": 0.72, "ab": 0.0}, {"d": "2026-01-19", "acc": 0.3529, "n": 34, "lo": 0.1731, "hi": 0.587, "ab": 0.0}, {"d": "2026-01-26", "acc": 0.4231, "n": 26, "lo": 0.204, "hi": 0.6772, "ab": 0.0769}, {"d": "2026-02-02", "acc": 0.4062, "n": 32, "lo": 0.2075, "hi": 0.6413, "ab": 0.0}, {"d": "2026-02-09", "acc": 0.6333, "n": 30, "lo": 0.3869, "hi": 0.8254, "ab": 0.0}, {"d": "2026-02-16", "acc": 0.3333, "n": 18, "lo": 0.1206, "hi": 0.6458, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.3929, "n": 28, "lo": 0.1881, "hi": 0.6437, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.3824, "n": 34, "lo": 0.1943, "hi": 0.6138, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.381, "n": 42, "lo": 0.2075, "hi": 0.5912, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.3929, "n": 28, "lo": 0.1881, "hi": 0.6437, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.55, "n": 20, "lo": 0.2737, "hi": 0.7986, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.6562, "n": 32, "lo": 0.4148, "hi": 0.8372, "ab": 0.0}, {"d": "2026-04-06", "acc": 0.5294, "n": 34, "lo": 0.3096, "hi": 0.7384, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.4118, "n": 34, "lo": 0.2161, "hi": 0.6399, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.6, "n": 30, "lo": 0.3575, "hi": 0.8018, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.2895, "n": 38, "lo": 0.1356, "hi": 0.5142, "ab": 0.0}, {"d": "2026-05-04", "acc": 0.3824, "n": 34, "lo": 0.1943, "hi": 0.6138, "ab": 0.0588}, {"d": "2026-05-11", "acc": 0.3235, "n": 34, "lo": 0.1526, "hi": 0.5595, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.3462, "n": 26, "lo": 0.1513, "hi": 0.6112, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.3929, "n": 28, "lo": 0.1881, "hi": 0.6437, "ab": 0.0714}, {"d": "2026-06-01", "acc": 0.3214, "n": 28, "lo": 0.1398, "hi": 0.58, "ab": 0.0}, {"d": "2026-06-08", "acc": 0.4545, "n": 22, "lo": 0.2127, "hi": 0.7199, "ab": 0.0909}, {"d": "2026-06-15", "acc": 0.3846, "n": 26, "lo": 0.1771, "hi": 0.6448, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.3462, "n": 26, "lo": 0.1513, "hi": 0.6112, "ab": 0.0769}, {"d": "2026-06-29", "acc": 0.3889, "n": 36, "lo": 0.203, "hi": 0.6138, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.1765, "n": 34, "lo": 0.0619, "hi": 0.4103, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.1579, "n": 38, "lo": 0.0552, "hi": 0.3757, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.3929, "n": 28, "lo": 0.1881, "hi": 0.6437, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.125, "n": 16, "lo": 0.0224, "hi": 0.4709, "ab": 0.0}], "samples": [{"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 92", "ok": true, "abstain": false, "conf": 92}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 97", "ok": true, "abstain": false, "conf": 97}, {"band": "pre", "date": "2024-08-14", "stem": "In mid-August 2024, four Nordic countries agreed to set up a joint police centre aimed at stopping Swedish criminal gang networks from spreading across the region. In which city was the hub established?", "correct_opt": "Stockholm", "model_opt": "Copenhagen", "raw": "E, 70", "ok": false, "abstain": false, "conf": 70}, {"band": "mid", "date": "2025-11-17", "stem": "In mid-November 2025, Cambridge University Press and Assessment revealed its pick for word of the year. Which word did it choose?", "correct_opt": "parasocial", "model_opt": "parasocial", "raw": "D, 90", "ok": true, "abstain": false, "conf": 90}, {"band": "mid", "date": "2025-11-17", "stem": "In mid-November 2025, the Philippine executive secretary and the budget secretary both stepped down amid the flood-control corruption scandal. Which cabinet member was moved into the vacated executive secretary post?", "correct_opt": "Finance secretary Ralph Recto", "model_opt": "Interior secretary Jonvic Remulla", "raw": "A, 70", "ok": false, "abstain": false, "conf": 70}, {"band": "mid", "date": "2025-11-17", "stem": "In mid-November 2025, dozens of schools across Australia and New Zealand shut their doors while officials investigated recalled children's play sand found to contain asbestos in some samples. Which retail chains had sol…", "correct_opt": "Kmart and Target", "model_opt": "Kmart and Target", "raw": "G, 70", "ok": true, "abstain": false, "conf": 70}, {"band": "mid", "date": "2025-11-18", "stem": "In mid-November 2025, Spanish police working with the U.S. Drug Enforcement Administration broke up a Mexican cartel's logistics hub in Spain and arrested 20 people, among them two members of an Italian criminal organis…", "correct_opt": "the Camorra", "model_opt": "the 'Ndrangheta", "raw": "C, 70", "ok": false, "abstain": false, "conf": 70}, {"band": "post", "date": "2026-01-19", "stem": "In mid-January 2026, amid the offensive in northeastern Syria, the SDF said it had lost control of a facility holding thousands of Islamic State detainees after it was attacked, with video appearing to show militants fl…", "correct_opt": "Al-Shaddadah", "model_opt": "Al-Hasakah", "raw": "A, 70", "ok": false, "abstain": false, "conf": 70}, {"band": "post", "date": "2026-01-19", "stem": "In mid-January 2026, the College Football Playoff title game ended 27–21. Which team lifted the national championship trophy?", "correct_opt": "Indiana Hoosiers", "model_opt": "Ohio State Buckeyes", "raw": "C, 35 This game occurred after my knowledge cutoff, so I can't know the actual result. Ba…", "ok": false, "abstain": false, "conf": 35}, {"band": "post", "date": "2026-01-19", "stem": "In mid-January 2026, a truck collided with a bus carrying pupils in South Africa's Gauteng province, killing thirteen children in the initial toll. Near which town did the crash occur?", "correct_opt": "Vanderbijlpark", "model_opt": "Carletonville", "raw": "F, 25", "ok": false, "abstain": false, "conf": 25}]}, {"id": "claude-sonnet-5", "name": "Claude Sonnet 5", "provider": "anthropic", "family": "sonnet", "release": "2026-06-30", "pub_rel": "2026-01", "pub_train": "2026-01", "pub_rel_d": "2026-01-31", "pub_train_d": "2026-01-31", "fit": {"midpoint": "2025-12-05", "width": 11.41, "decay_start": "2025-11-08", "floor_reach": "2026-01-11", "floor": 0.2552, "plateau": 0.5897, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2025-11-17", "2026-01-05"]}, "weekly": [{"d": "2024-07-29", "acc": 0.625, "n": 16, "lo": 0.3057, "hi": 0.8632, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.5, "n": 26, "lo": 0.2612, "hi": 0.7388, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.4583, "n": 24, "lo": 0.2229, "hi": 0.714, "ab": 0.0}, {"d": "2024-08-19", "acc": 0.3846, "n": 26, "lo": 0.1771, "hi": 0.6448, "ab": 0.0}, {"d": "2024-08-26", "acc": 0.6818, "n": 22, "lo": 0.3932, "hi": 0.8764, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.5625, "n": 32, "lo": 0.3318, "hi": 0.769, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.4615, "n": 26, "lo": 0.2321, "hi": 0.7086, "ab": 0.0}, {"d": "2024-09-16", "acc": 0.4643, "n": 28, "lo": 0.2404, "hi": 0.7035, "ab": 0.0}, {"d": "2024-09-23", "acc": 0.5909, "n": 22, "lo": 0.3161, "hi": 0.8186, "ab": 0.0}, {"d": "2024-09-30", "acc": 0.375, "n": 24, "lo": 0.165, "hi": 0.6456, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.6429, "n": 14, "lo": 0.3026, "hi": 0.8819, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.7273, "n": 22, "lo": 0.4343, "hi": 0.9025, "ab": 0.0}, {"d": "2024-10-21", "acc": 0.4375, "n": 16, "lo": 0.1745, "hi": 0.7411, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.4583, "n": 24, "lo": 0.2229, "hi": 0.714, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.3636, "n": 22, "lo": 0.1517, "hi": 0.6462, "ab": 0.0}, {"d": "2024-11-11", "acc": 0.5417, "n": 24, "lo": 0.286, "hi": 0.7771, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.7778, "n": 18, "lo": 0.4526, "hi": 0.9368, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.6176, "n": 34, "lo": 0.3862, "hi": 0.8057, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.5455, "n": 22, "lo": 0.2801, "hi": 0.7873, "ab": 0.0}, {"d": "2024-12-09", "acc": 0.6333, "n": 30, "lo": 0.3869, "hi": 0.8254, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.6842, "n": 38, "lo": 0.4601, "hi": 0.8464, "ab": 0.0}, {"d": "2024-12-23", "acc": 0.5, "n": 40, "lo": 0.2993, "hi": 0.7007, "ab": 0.0}, {"d": "2024-12-30", "acc": 0.6, "n": 10, "lo": 0.2307, "hi": 0.8824, "ab": 0.0}, {"d": "2025-01-06", "acc": 0.7778, "n": 18, "lo": 0.4526, "hi": 0.9368, "ab": 0.0}, {"d": "2025-01-13", "acc": 0.6875, "n": 16, "lo": 0.3558, "hi": 0.8976, "ab": 0.0}, {"d": "2025-01-20", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2025-01-27", "acc": 0.5333, "n": 30, "lo": 0.3012, "hi": 0.7519, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.4286, "n": 28, "lo": 0.2138, "hi": 0.6741, "ab": 0.0}, {"d": "2025-02-10", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2025-02-17", "acc": 0.6333, "n": 30, "lo": 0.3869, "hi": 0.8254, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.6429, "n": 28, "lo": 0.3876, "hi": 0.8366, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.625, "n": 32, "lo": 0.3864, "hi": 0.8152, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.6944, "n": 36, "lo": 0.4641, "hi": 0.8564, "ab": 0.0}, {"d": "2025-03-17", "acc": 0.6154, "n": 26, "lo": 0.3552, "hi": 0.8229, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.5, "n": 20, "lo": 0.2366, "hi": 0.7634, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.7143, "n": 28, "lo": 0.4535, "hi": 0.8828, "ab": 0.0}, {"d": "2025-04-07", "acc": 0.5556, "n": 36, "lo": 0.3372, "hi": 0.7544, "ab": 0.0}, {"d": "2025-04-14", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.4706, "n": 34, "lo": 0.2616, "hi": 0.6904, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.8214, "n": 28, "lo": 0.5615, "hi": 0.9429, "ab": 0.0}, {"d": "2025-05-05", "acc": 0.5714, "n": 28, "lo": 0.3259, "hi": 0.7862, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.5263, "n": 38, "lo": 0.3171, "hi": 0.7267, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.6333, "n": 30, "lo": 0.3869, "hi": 0.8254, "ab": 0.0}, {"d": "2025-05-26", "acc": 0.7083, "n": 24, "lo": 0.4284, "hi": 0.8873, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.5294, "n": 34, "lo": 0.3096, "hi": 0.7384, "ab": 0.0}, {"d": "2025-06-09", "acc": 0.7083, "n": 24, "lo": 0.4284, "hi": 0.8873, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.5667, "n": 30, "lo": 0.3289, "hi": 0.7772, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.5588, "n": 34, "lo": 0.3345, "hi": 0.7614, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.6842, "n": 38, "lo": 0.4601, "hi": 0.8464, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.8333, "n": 24, "lo": 0.552, "hi": 0.953, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.4091, "n": 22, "lo": 0.1814, "hi": 0.6839, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.6333, "n": 30, "lo": 0.3869, "hi": 0.8254, "ab": 0.0}, {"d": "2025-07-28", "acc": 0.6, "n": 30, "lo": 0.3575, "hi": 0.8018, "ab": 0.0}, {"d": "2025-08-04", "acc": 0.6111, "n": 18, "lo": 0.3092, "hi": 0.8466, "ab": 0.0}, {"d": "2025-08-11", "acc": 0.3438, "n": 32, "lo": 0.1628, "hi": 0.5852, "ab": 0.0}, {"d": "2025-08-18", "acc": 0.6579, "n": 38, "lo": 0.435, "hi": 0.8277, "ab": 0.0}, {"d": "2025-08-25", "acc": 0.619, "n": 42, "lo": 0.4088, "hi": 0.7925, "ab": 0.0}, {"d": "2025-09-01", "acc": 0.7, "n": 20, "lo": 0.3968, "hi": 0.8922, "ab": 0.0}, {"d": "2025-09-08", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-09-15", "acc": 0.5455, "n": 22, "lo": 0.2801, "hi": 0.7873, "ab": 0.0}, {"d": "2025-09-22", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2025-09-29", "acc": 0.5909, "n": 22, "lo": 0.3161, "hi": 0.8186, "ab": 0.0}, {"d": "2025-10-06", "acc": 0.5, "n": 34, "lo": 0.2853, "hi": 0.7147, "ab": 0.0}, {"d": "2025-10-13", "acc": 0.5833, "n": 36, "lo": 0.3614, "hi": 0.776, "ab": 0.0}, {"d": "2025-10-20", "acc": 0.425, "n": 40, "lo": 0.2383, "hi": 0.6359, "ab": 0.0}, {"d": "2025-10-27", "acc": 0.5476, "n": 42, "lo": 0.3444, "hi": 0.7361, "ab": 0.0}, {"d": "2025-11-03", "acc": 0.4231, "n": 26, "lo": 0.204, "hi": 0.6772, "ab": 0.0}, {"d": "2025-11-10", "acc": 0.6429, "n": 28, "lo": 0.3876, "hi": 0.8366, "ab": 0.0}, {"d": "2025-11-17", "acc": 0.4545, "n": 22, "lo": 0.2127, "hi": 0.7199, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.7, "n": 30, "lo": 0.4483, "hi": 0.8701, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.3333, "n": 30, "lo": 0.1518, "hi": 0.5829, "ab": 0.0}, {"d": "2025-12-08", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.4211, "n": 38, "lo": 0.2314, "hi": 0.6372, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.0}, {"d": "2025-12-29", "acc": 0.3333, "n": 24, "lo": 0.1381, "hi": 0.6094, "ab": 0.0}, {"d": "2026-01-05", "acc": 0.3462, "n": 26, "lo": 0.1513, "hi": 0.6112, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.125, "n": 32, "lo": 0.035, "hi": 0.3602, "ab": 0.0}, {"d": "2026-01-19", "acc": 0.1471, "n": 34, "lo": 0.0467, "hi": 0.3775, "ab": 0.0}, {"d": "2026-01-26", "acc": 0.3846, "n": 26, "lo": 0.1771, "hi": 0.6448, "ab": 0.0}, {"d": "2026-02-02", "acc": 0.0312, "n": 32, "lo": 0.0033, "hi": 0.2407, "ab": 0.0}, {"d": "2026-02-09", "acc": 0.3, "n": 30, "lo": 0.1299, "hi": 0.5517, "ab": 0.0333}, {"d": "2026-02-16", "acc": 0.3333, "n": 18, "lo": 0.1206, "hi": 0.6458, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.3214, "n": 28, "lo": 0.1398, "hi": 0.58, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.4412, "n": 34, "lo": 0.2386, "hi": 0.6655, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.2381, "n": 42, "lo": 0.1063, "hi": 0.4509, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.3214, "n": 28, "lo": 0.1398, "hi": 0.58, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.5, "n": 20, "lo": 0.2366, "hi": 0.7634, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.5938, "n": 32, "lo": 0.3587, "hi": 0.7925, "ab": 0.0}, {"d": "2026-04-06", "acc": 0.4412, "n": 34, "lo": 0.2386, "hi": 0.6655, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.3235, "n": 34, "lo": 0.1526, "hi": 0.5595, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.3, "n": 30, "lo": 0.1299, "hi": 0.5517, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.3158, "n": 38, "lo": 0.1536, "hi": 0.5399, "ab": 0.0}, {"d": "2026-05-04", "acc": 0.2353, "n": 34, "lo": 0.0955, "hi": 0.4726, "ab": 0.0}, {"d": "2026-05-11", "acc": 0.1471, "n": 34, "lo": 0.0467, "hi": 0.3775, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.2308, "n": 26, "lo": 0.0818, "hi": 0.5026, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.2857, "n": 28, "lo": 0.1172, "hi": 0.5465, "ab": 0.0357}, {"d": "2026-06-01", "acc": 0.1429, "n": 28, "lo": 0.0401, "hi": 0.3994, "ab": 0.0}, {"d": "2026-06-08", "acc": 0.1818, "n": 22, "lo": 0.0514, "hi": 0.477, "ab": 0.0}, {"d": "2026-06-15", "acc": 0.0769, "n": 26, "lo": 0.0137, "hi": 0.3331, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.2308, "n": 26, "lo": 0.0818, "hi": 0.5026, "ab": 0.0}, {"d": "2026-06-29", "acc": 0.2222, "n": 36, "lo": 0.09, "hi": 0.4522, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.0588, "n": 34, "lo": 0.0105, "hi": 0.2698, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.0789, "n": 38, "lo": 0.0184, "hi": 0.2811, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.1875, "n": 16, "lo": 0.0445, "hi": 0.5332, "ab": 0.0}], "samples": [{"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 85", "ok": true, "abstain": false, "conf": 85}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 85", "ok": true, "abstain": false, "conf": 85}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, a South American government became the third country in the world to recognise Edmundo González as the rightful winner of Venezuela's disputed presidential election. Which country was it?", "correct_opt": "Argentina", "model_opt": "Paraguay", "raw": "B, 55", "ok": false, "abstain": false, "conf": 55}, {"band": "mid", "date": "2025-11-10", "stem": "In early November 2025, Nigeria's anti-corruption commission sought the arrest of a former petroleum minister over alleged fraud involving nearly $15 million earmarked for a refinery project. Who was the official named?", "correct_opt": "Timipre Sylva", "model_opt": "Timipre Sylva", "raw": "G, 62", "ok": true, "abstain": false, "conf": 62}, {"band": "mid", "date": "2025-11-10", "stem": "In early November 2025, Ecuador's president moved 300 high-risk inmates, among them a jailed former vice president, into a newly used maximum-security facility. Which prison received them?", "correct_opt": "El Encuentro in Santa Elena", "model_opt": "La Roca in Guayaquil", "raw": "H, 62", "ok": false, "abstain": false, "conf": 62}, {"band": "mid", "date": "2025-11-10", "stem": "In early November 2025, Thailand froze its implementation of a recently signed accord with Cambodia after two of its soldiers were hurt by a land mine. In which Thai province did that blast occur?", "correct_opt": "Sisaket", "model_opt": "Sisaket", "raw": "B, 55", "ok": true, "abstain": false, "conf": 55}, {"band": "mid", "date": "2025-11-12", "stem": "In mid-November 2025, six people died in a crush at a military recruitment drive in a West African capital. In which city did it happen?", "correct_opt": "Accra", "model_opt": "Conakry", "raw": "F, 55", "ok": false, "abstain": false, "conf": 55}, {"band": "post", "date": "2026-01-12", "stem": "In mid-January 2026, a former prime minister of Georgia admitted guilt on large-scale money-laundering charges after investigators seized millions of dollars in cash during searches of his home. Who was this former prim…", "correct_opt": "Irakli Garibashvili", "model_opt": "Giorgi Kvirikashvili", "raw": "B, 55", "ok": false, "abstain": false, "conf": 55}, {"band": "post", "date": "2026-01-12", "stem": "In mid-January 2026, a former leader of a Corsican nationalist armed movement was shot dead by a gunman while attending his mother's funeral. What was the victim's name?", "correct_opt": "Alain Orsoni", "model_opt": "Charles Pieri", "raw": "D, 55", "ok": false, "abstain": false, "conf": 55}, {"band": "post", "date": "2026-01-12", "stem": "In mid-January 2026, Spanish police working with British and American agencies intercepted a cargo ship off the Canary Islands in the country's biggest ever maritime drug haul. What cargo was the cocaine concealed withi…", "correct_opt": "Salt", "model_opt": "Soybean meal", "raw": "B, 45", "ok": false, "abstain": false, "conf": 45}]}, {"id": "claude-opus-5", "name": "Claude Opus 5", "provider": "anthropic", "family": "opus", "release": "2026-07-24", "pub_rel": "2026-05", "pub_train": "2026-05", "pub_rel_d": "2026-05-31", "pub_train_d": "2026-05-31", "fit": {"midpoint": "2025-12-04", "width": 13.32, "decay_start": "2025-11-04", "floor_reach": "2026-01-13", "floor": 0.3587, "plateau": 0.7368, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2025-11-23", "2026-01-17"]}, "weekly": [{"d": "2024-07-29", "acc": 0.75, "n": 16, "lo": 0.4093, "hi": 0.9285, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.5385, "n": 26, "lo": 0.2914, "hi": 0.7679, "ab": 0.1538}, {"d": "2024-08-12", "acc": 0.6667, "n": 24, "lo": 0.3906, "hi": 0.8619, "ab": 0.0417}, {"d": "2024-08-19", "acc": 0.7308, "n": 26, "lo": 0.4598, "hi": 0.8964, "ab": 0.0385}, {"d": "2024-08-26", "acc": 0.7727, "n": 22, "lo": 0.4775, "hi": 0.9267, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.6562, "n": 32, "lo": 0.4148, "hi": 0.8372, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.7308, "n": 26, "lo": 0.4598, "hi": 0.8964, "ab": 0.0769}, {"d": "2024-09-16", "acc": 0.6429, "n": 28, "lo": 0.3876, "hi": 0.8366, "ab": 0.0714}, {"d": "2024-09-23", "acc": 0.6818, "n": 22, "lo": 0.3932, "hi": 0.8764, "ab": 0.0}, {"d": "2024-09-30", "acc": 0.5833, "n": 24, "lo": 0.3195, "hi": 0.8067, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.7857, "n": 14, "lo": 0.4201, "hi": 0.9489, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.5909, "n": 22, "lo": 0.3161, "hi": 0.8186, "ab": 0.0}, {"d": "2024-10-21", "acc": 0.625, "n": 16, "lo": 0.3057, "hi": 0.8632, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.5417, "n": 24, "lo": 0.286, "hi": 0.7771, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.5909, "n": 22, "lo": 0.3161, "hi": 0.8186, "ab": 0.0}, {"d": "2024-11-11", "acc": 0.75, "n": 24, "lo": 0.4677, "hi": 0.9111, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.7778, "n": 18, "lo": 0.4526, "hi": 0.9368, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.6765, "n": 34, "lo": 0.4405, "hi": 0.8474, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.5, "n": 22, "lo": 0.2456, "hi": 0.7544, "ab": 0.0}, {"d": "2024-12-09", "acc": 0.8, "n": 30, "lo": 0.5481, "hi": 0.9295, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.8421, "n": 38, "lo": 0.6243, "hi": 0.9448, "ab": 0.0263}, {"d": "2024-12-23", "acc": 0.7, "n": 40, "lo": 0.481, "hi": 0.8545, "ab": 0.0}, {"d": "2024-12-30", "acc": 0.8, "n": 10, "lo": 0.3755, "hi": 0.9638, "ab": 0.0}, {"d": "2025-01-06", "acc": 0.8889, "n": 18, "lo": 0.565, "hi": 0.9801, "ab": 0.0}, {"d": "2025-01-13", "acc": 1.0, "n": 16, "lo": 0.6756, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-20", "acc": 0.9545, "n": 22, "lo": 0.6785, "hi": 0.9952, "ab": 0.0}, {"d": "2025-01-27", "acc": 0.7, "n": 30, "lo": 0.4483, "hi": 0.8701, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.7143, "n": 28, "lo": 0.4535, "hi": 0.8828, "ab": 0.0}, {"d": "2025-02-10", "acc": 0.9545, "n": 22, "lo": 0.6785, "hi": 0.9952, "ab": 0.0}, {"d": "2025-02-17", "acc": 0.8333, "n": 30, "lo": 0.5839, "hi": 0.9469, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.8214, "n": 28, "lo": 0.5615, "hi": 0.9429, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.875, "n": 32, "lo": 0.6398, "hi": 0.965, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.7778, "n": 36, "lo": 0.5478, "hi": 0.91, "ab": 0.0}, {"d": "2025-03-17", "acc": 0.8846, "n": 26, "lo": 0.6209, "hi": 0.9729, "ab": 0.0385}, {"d": "2025-03-24", "acc": 0.5, "n": 20, "lo": 0.2366, "hi": 0.7634, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.8571, "n": 28, "lo": 0.6006, "hi": 0.9599, "ab": 0.0714}, {"d": "2025-04-07", "acc": 0.75, "n": 36, "lo": 0.5192, "hi": 0.8929, "ab": 0.0278}, {"d": "2025-04-14", "acc": 0.5455, "n": 22, "lo": 0.2801, "hi": 0.7873, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.7353, "n": 34, "lo": 0.4976, "hi": 0.8862, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.7857, "n": 28, "lo": 0.5241, "hi": 0.9243, "ab": 0.0357}, {"d": "2025-05-05", "acc": 0.7143, "n": 28, "lo": 0.4535, "hi": 0.8828, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.4737, "n": 38, "lo": 0.2733, "hi": 0.6829, "ab": 0.0526}, {"d": "2025-05-19", "acc": 0.8667, "n": 30, "lo": 0.6212, "hi": 0.9626, "ab": 0.0}, {"d": "2025-05-26", "acc": 0.875, "n": 24, "lo": 0.5975, "hi": 0.9706, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.6471, "n": 34, "lo": 0.413, "hi": 0.8269, "ab": 0.0}, {"d": "2025-06-09", "acc": 0.75, "n": 24, "lo": 0.4677, "hi": 0.9111, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.7333, "n": 30, "lo": 0.4805, "hi": 0.891, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.7059, "n": 34, "lo": 0.4687, "hi": 0.8672, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.6316, "n": 38, "lo": 0.4104, "hi": 0.8085, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.75, "n": 24, "lo": 0.4677, "hi": 0.9111, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.6818, "n": 22, "lo": 0.3932, "hi": 0.8764, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.9667, "n": 30, "lo": 0.7465, "hi": 0.9965, "ab": 0.0}, {"d": "2025-07-28", "acc": 0.7, "n": 30, "lo": 0.4483, "hi": 0.8701, "ab": 0.0}, {"d": "2025-08-04", "acc": 0.6667, "n": 18, "lo": 0.3542, "hi": 0.8794, "ab": 0.0}, {"d": "2025-08-11", "acc": 0.7188, "n": 32, "lo": 0.4741, "hi": 0.8787, "ab": 0.0938}, {"d": "2025-08-18", "acc": 0.8421, "n": 38, "lo": 0.6243, "hi": 0.9448, "ab": 0.0263}, {"d": "2025-08-25", "acc": 0.8333, "n": 42, "lo": 0.6264, "hi": 0.9371, "ab": 0.0}, {"d": "2025-09-01", "acc": 0.75, "n": 20, "lo": 0.4422, "hi": 0.9191, "ab": 0.0}, {"d": "2025-09-08", "acc": 0.9643, "n": 28, "lo": 0.7324, "hi": 0.9963, "ab": 0.0}, {"d": "2025-09-15", "acc": 0.6818, "n": 22, "lo": 0.3932, "hi": 0.8764, "ab": 0.0455}, {"d": "2025-09-22", "acc": 0.9545, "n": 22, "lo": 0.6785, "hi": 0.9952, "ab": 0.0}, {"d": "2025-09-29", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2025-10-06", "acc": 0.6176, "n": 34, "lo": 0.3862, "hi": 0.8057, "ab": 0.0294}, {"d": "2025-10-13", "acc": 0.6944, "n": 36, "lo": 0.4641, "hi": 0.8564, "ab": 0.0556}, {"d": "2025-10-20", "acc": 0.6, "n": 40, "lo": 0.3866, "hi": 0.7812, "ab": 0.0}, {"d": "2025-10-27", "acc": 0.6905, "n": 42, "lo": 0.4769, "hi": 0.8452, "ab": 0.0238}, {"d": "2025-11-03", "acc": 0.7692, "n": 26, "lo": 0.4974, "hi": 0.9182, "ab": 0.0}, {"d": "2025-11-10", "acc": 0.6071, "n": 28, "lo": 0.3563, "hi": 0.8119, "ab": 0.0}, {"d": "2025-11-17", "acc": 0.6818, "n": 22, "lo": 0.3932, "hi": 0.8764, "ab": 0.0455}, {"d": "2025-11-24", "acc": 0.7667, "n": 30, "lo": 0.5137, "hi": 0.9109, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.4667, "n": 30, "lo": 0.2481, "hi": 0.6988, "ab": 0.0333}, {"d": "2025-12-08", "acc": 0.5, "n": 22, "lo": 0.2456, "hi": 0.7544, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.4211, "n": 38, "lo": 0.2314, "hi": 0.6372, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.4545, "n": 22, "lo": 0.2127, "hi": 0.7199, "ab": 0.0455}, {"d": "2025-12-29", "acc": 0.4583, "n": 24, "lo": 0.2229, "hi": 0.714, "ab": 0.0}, {"d": "2026-01-05", "acc": 0.3077, "n": 26, "lo": 0.1268, "hi": 0.5763, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.3438, "n": 32, "lo": 0.1628, "hi": 0.5852, "ab": 0.0312}, {"d": "2026-01-19", "acc": 0.3824, "n": 34, "lo": 0.1943, "hi": 0.6138, "ab": 0.0294}, {"d": "2026-01-26", "acc": 0.5, "n": 26, "lo": 0.2612, "hi": 0.7388, "ab": 0.0}, {"d": "2026-02-02", "acc": 0.25, "n": 32, "lo": 0.1018, "hi": 0.495, "ab": 0.0625}, {"d": "2026-02-09", "acc": 0.4667, "n": 30, "lo": 0.2481, "hi": 0.6988, "ab": 0.0}, {"d": "2026-02-16", "acc": 0.3333, "n": 18, "lo": 0.1206, "hi": 0.6458, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.3235, "n": 34, "lo": 0.1526, "hi": 0.5595, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.4286, "n": 42, "lo": 0.2447, "hi": 0.6345, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.3214, "n": 28, "lo": 0.1398, "hi": 0.58, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.45, "n": 20, "lo": 0.2014, "hi": 0.7263, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.625, "n": 32, "lo": 0.3864, "hi": 0.8152, "ab": 0.0}, {"d": "2026-04-06", "acc": 0.5588, "n": 34, "lo": 0.3345, "hi": 0.7614, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.5882, "n": 34, "lo": 0.3601, "hi": 0.7839, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.3667, "n": 30, "lo": 0.1746, "hi": 0.6131, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.2105, "n": 38, "lo": 0.0851, "hi": 0.4333, "ab": 0.0526}, {"d": "2026-05-04", "acc": 0.4118, "n": 34, "lo": 0.2161, "hi": 0.6399, "ab": 0.0294}, {"d": "2026-05-11", "acc": 0.3824, "n": 34, "lo": 0.1943, "hi": 0.6138, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.2692, "n": 26, "lo": 0.1036, "hi": 0.5402, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.4286, "n": 28, "lo": 0.2138, "hi": 0.6741, "ab": 0.0}, {"d": "2026-06-01", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.0357}, {"d": "2026-06-08", "acc": 0.4091, "n": 22, "lo": 0.1814, "hi": 0.6839, "ab": 0.0}, {"d": "2026-06-15", "acc": 0.4231, "n": 26, "lo": 0.204, "hi": 0.6772, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.3077, "n": 26, "lo": 0.1268, "hi": 0.5763, "ab": 0.0}, {"d": "2026-06-29", "acc": 0.25, "n": 36, "lo": 0.1071, "hi": 0.4808, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.1765, "n": 34, "lo": 0.0619, "hi": 0.4103, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.1579, "n": 38, "lo": 0.0552, "hi": 0.3757, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.1875, "n": 16, "lo": 0.0445, "hi": 0.5332, "ab": 0.0}], "samples": [{"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 97", "ok": true, "abstain": false, "conf": 97}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 96", "ok": true, "abstain": false, "conf": 96}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, a South American government became the third country in the world to recognise Edmundo González as the rightful winner of Venezuela's disputed presidential election. Which country was it?", "correct_opt": "Argentina", "model_opt": "Uruguay", "raw": "<answer>C<\/answer>, 45", "ok": false, "abstain": false, "conf": 45}, {"band": "mid", "date": "2025-11-05", "stem": "In early November 2025, the Czech Chamber of Deputies chose a new speaker, the first big post filled under Andrej Babiš's incoming coalition. Who was elected to the role?", "correct_opt": "Tomio Okamura", "model_opt": "Tomio Okamura", "raw": "H, 95", "ok": true, "abstain": false, "conf": 95}, {"band": "mid", "date": "2025-11-05", "stem": "In early November 2025, a German court jailed a palliative care nurse for life after finding he had killed ten patients with lethal injections at a hospital in the country's west. In which city did the court sit?", "correct_opt": "Aachen", "model_opt": "Essen", "raw": "A, 85", "ok": false, "abstain": false, "conf": 85}, {"band": "mid", "date": "2025-11-05", "stem": "In early November 2025, U.S. health authorities linked a multi-state listeria outbreak that killed six people to a recalled category of supermarket food. Which product was blamed?", "correct_opt": "prepared pasta meals", "model_opt": "prepared pasta meals", "raw": "**F, 85** Contextual answer: In November", "ok": true, "abstain": false, "conf": 85}, {"band": "mid", "date": "2025-11-06", "stem": "In early November 2025, a court in Russian-held Donetsk jailed two foreign nationals for having fought alongside Ukrainian forces; they had been seized while travelling home through Venezuela. What was their nationality?", "correct_opt": "Colombian", "model_opt": "Cuban", "raw": "D, 72", "ok": false, "abstain": false, "conf": 72}, {"band": "post", "date": "2026-01-13", "stem": "In mid-January 2026, an Australian literary festival scrapped its entire 2026 programme after its director quit and roughly 180 authors pulled out over the board's move to rescind an invitation to a Palestinian Australi…", "correct_opt": "Adelaide Writers' Week", "model_opt": "Byron Writers Festival", "raw": "A, 68", "ok": false, "abstain": false, "conf": 68}, {"band": "post", "date": "2026-01-13", "stem": "In mid-January 2026, the United Arab Emirates concluded a free trade agreement with a Southeast Asian nation. Which country signed the deal with the UAE?", "correct_opt": "The Philippines", "model_opt": "Thailand", "raw": "E, 45", "ok": false, "abstain": false, "conf": 45}, {"band": "post", "date": "2026-01-14", "stem": "In mid-January 2026, two Swiss cantons prohibited pyrotechnics inside enclosed venues after a nightclub blaze in an Alpine resort killed 40 people. In which resort did the deadly fire occur?", "correct_opt": "Crans-Montana", "model_opt": "St. Moritz", "raw": "A, 20 I don't have knowledge of this event", "ok": false, "abstain": false, "conf": 20}]}, {"id": "gpt-5", "name": "GPT-5", "provider": "openai", "family": "gpt-5", "release": "2025-08-07", "pub_rel": "2024-09", "pub_train": "2024-09", "pub_rel_d": "2024-09-30", "pub_train_d": "2024-09-30", "fit": {"midpoint": "2024-06-28", "width": 21.59, "decay_start": "2024-04-29", "floor_reach": "2024-10-07", "floor": 0.1567, "plateau": 1.0, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2024-06-28", "2024-09-03"]}, "weekly": [{"d": "2024-07-29", "acc": 0.3125, "n": 16, "lo": 0.1024, "hi": 0.6442, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.2692, "n": 26, "lo": 0.1036, "hi": 0.5402, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.125, "n": 24, "lo": 0.0294, "hi": 0.4025, "ab": 0.0}, {"d": "2024-08-19", "acc": 0.2692, "n": 26, "lo": 0.1036, "hi": 0.5402, "ab": 0.0}, {"d": "2024-08-26", "acc": 0.2273, "n": 22, "lo": 0.0733, "hi": 0.5225, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.25, "n": 32, "lo": 0.1018, "hi": 0.495, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.1154, "n": 26, "lo": 0.0271, "hi": 0.3791, "ab": 0.0}, {"d": "2024-09-16", "acc": 0.1429, "n": 28, "lo": 0.0401, "hi": 0.3994, "ab": 0.0}, {"d": "2024-09-23", "acc": 0.3636, "n": 22, "lo": 0.1517, "hi": 0.6462, "ab": 0.0}, {"d": "2024-09-30", "acc": 0.0833, "n": 24, "lo": 0.0149, "hi": 0.3539, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.1429, "n": 14, "lo": 0.0257, "hi": 0.5131, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.0}, {"d": "2024-10-21", "acc": 0.125, "n": 16, "lo": 0.0224, "hi": 0.4709, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.2083, "n": 24, "lo": 0.0669, "hi": 0.4912, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.0}, {"d": "2024-11-11", "acc": 0.25, "n": 24, "lo": 0.0889, "hi": 0.5323, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.0556, "n": 18, "lo": 0.0058, "hi": 0.3712, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.0882, "n": 34, "lo": 0.0206, "hi": 0.3076, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.0455, "n": 22, "lo": 0.0048, "hi": 0.3215, "ab": 0.0}, {"d": "2024-12-09", "acc": 0.0667, "n": 30, "lo": 0.0119, "hi": 0.2982, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.1053, "n": 38, "lo": 0.0294, "hi": 0.3139, "ab": 0.0}, {"d": "2024-12-23", "acc": 0.175, "n": 40, "lo": 0.0661, "hi": 0.3886, "ab": 0.0}, {"d": "2024-12-30", "acc": 0.1, "n": 10, "lo": 0.0105, "hi": 0.5371, "ab": 0.0}, {"d": "2025-01-06", "acc": 0.2222, "n": 18, "lo": 0.0632, "hi": 0.5474, "ab": 0.0}, {"d": "2025-01-13", "acc": 0.125, "n": 16, "lo": 0.0224, "hi": 0.4709, "ab": 0.0}, {"d": "2025-01-20", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.0}, {"d": "2025-01-27", "acc": 0.3667, "n": 30, "lo": 0.1746, "hi": 0.6131, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.0}, {"d": "2025-02-10", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.0}, {"d": "2025-02-17", "acc": 0.1667, "n": 30, "lo": 0.0531, "hi": 0.4161, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.2188, "n": 32, "lo": 0.0833, "hi": 0.4631, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.1111, "n": 36, "lo": 0.031, "hi": 0.328, "ab": 0.0}, {"d": "2025-03-17", "acc": 0.1538, "n": 26, "lo": 0.0433, "hi": 0.4224, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.25, "n": 20, "lo": 0.0809, "hi": 0.5578, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.0}, {"d": "2025-04-07", "acc": 0.1667, "n": 36, "lo": 0.0584, "hi": 0.3922, "ab": 0.0}, {"d": "2025-04-14", "acc": 0.0455, "n": 22, "lo": 0.0048, "hi": 0.3215, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.2059, "n": 34, "lo": 0.0782, "hi": 0.442, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.0}, {"d": "2025-05-05", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.2632, "n": 38, "lo": 0.1181, "hi": 0.4879, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.2, "n": 30, "lo": 0.0705, "hi": 0.4519, "ab": 0.0}, {"d": "2025-05-26", "acc": 0.125, "n": 24, "lo": 0.0294, "hi": 0.4025, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.0294, "n": 34, "lo": 0.0031, "hi": 0.2292, "ab": 0.0}, {"d": "2025-06-09", "acc": 0.125, "n": 24, "lo": 0.0294, "hi": 0.4025, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.3333, "n": 30, "lo": 0.1518, "hi": 0.5829, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.1176, "n": 34, "lo": 0.0329, "hi": 0.3434, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.1579, "n": 38, "lo": 0.0552, "hi": 0.3757, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.125, "n": 24, "lo": 0.0294, "hi": 0.4025, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.3636, "n": 22, "lo": 0.1517, "hi": 0.6462, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.2667, "n": 30, "lo": 0.109, "hi": 0.5195, "ab": 0.0}, {"d": "2025-07-28", "acc": 0.2, "n": 30, "lo": 0.0705, "hi": 0.4519, "ab": 0.0}, {"d": "2025-08-04", "acc": 0.0556, "n": 18, "lo": 0.0058, "hi": 0.3712, "ab": 0.0}, {"d": "2025-08-11", "acc": 0.0312, "n": 32, "lo": 0.0033, "hi": 0.2407, "ab": 0.0}, {"d": "2025-08-18", "acc": 0.1053, "n": 38, "lo": 0.0294, "hi": 0.3139, "ab": 0.0}, {"d": "2025-08-25", "acc": 0.2143, "n": 42, "lo": 0.0912, "hi": 0.4258, "ab": 0.0}, {"d": "2025-09-01", "acc": 0.3, "n": 20, "lo": 0.1078, "hi": 0.6032, "ab": 0.0}, {"d": "2025-09-08", "acc": 0.0714, "n": 28, "lo": 0.0127, "hi": 0.3147, "ab": 0.0}, {"d": "2025-09-15", "acc": 0.0455, "n": 22, "lo": 0.0048, "hi": 0.3215, "ab": 0.0}, {"d": "2025-09-22", "acc": 0.1818, "n": 22, "lo": 0.0514, "hi": 0.477, "ab": 0.0}, {"d": "2025-09-29", "acc": 0.0909, "n": 22, "lo": 0.0162, "hi": 0.3774, "ab": 0.0}, {"d": "2025-10-06", "acc": 0.0294, "n": 34, "lo": 0.0031, "hi": 0.2292, "ab": 0.0}, {"d": "2025-10-13", "acc": 0.1389, "n": 36, "lo": 0.0441, "hi": 0.3607, "ab": 0.0}, {"d": "2025-10-20", "acc": 0.25, "n": 40, "lo": 0.1119, "hi": 0.4687, "ab": 0.0}, {"d": "2025-10-27", "acc": 0.2619, "n": 42, "lo": 0.1219, "hi": 0.4755, "ab": 0.0}, {"d": "2025-11-03", "acc": 0.1923, "n": 26, "lo": 0.0616, "hi": 0.4634, "ab": 0.0}, {"d": "2025-11-10", "acc": 0.1071, "n": 28, "lo": 0.0251, "hi": 0.3583, "ab": 0.0}, {"d": "2025-11-17", "acc": 0.0909, "n": 22, "lo": 0.0162, "hi": 0.3774, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.1333, "n": 30, "lo": 0.0374, "hi": 0.3788, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.0667, "n": 30, "lo": 0.0119, "hi": 0.2982, "ab": 0.0}, {"d": "2025-12-08", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.0789, "n": 38, "lo": 0.0184, "hi": 0.2811, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.0909, "n": 22, "lo": 0.0162, "hi": 0.3774, "ab": 0.0}, {"d": "2025-12-29", "acc": 0.1667, "n": 24, "lo": 0.047, "hi": 0.448, "ab": 0.0}, {"d": "2026-01-05", "acc": 0.1538, "n": 26, "lo": 0.0433, "hi": 0.4224, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.0312, "n": 32, "lo": 0.0033, "hi": 0.2407, "ab": 0.0}, {"d": "2026-01-19", "acc": 0.0588, "n": 34, "lo": 0.0105, "hi": 0.2698, "ab": 0.0}, {"d": "2026-01-26", "acc": 0.2692, "n": 26, "lo": 0.1036, "hi": 0.5402, "ab": 0.0}, {"d": "2026-02-02", "acc": 0.2188, "n": 32, "lo": 0.0833, "hi": 0.4631, "ab": 0.0}, {"d": "2026-02-09", "acc": 0.0667, "n": 30, "lo": 0.0119, "hi": 0.2982, "ab": 0.0}, {"d": "2026-02-16", "acc": 0.4444, "n": 18, "lo": 0.1888, "hi": 0.7334, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.1429, "n": 28, "lo": 0.0401, "hi": 0.3994, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.2941, "n": 34, "lo": 0.1328, "hi": 0.5313, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.0952, "n": 42, "lo": 0.0265, "hi": 0.2891, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.0714, "n": 28, "lo": 0.0127, "hi": 0.3147, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.15, "n": 20, "lo": 0.0354, "hi": 0.4589, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.1875, "n": 32, "lo": 0.0659, "hi": 0.4301, "ab": 0.0}, {"d": "2026-04-06", "acc": 0.1176, "n": 34, "lo": 0.0329, "hi": 0.3434, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.1176, "n": 34, "lo": 0.0329, "hi": 0.3434, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.1, "n": 30, "lo": 0.0234, "hi": 0.3397, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.0789, "n": 38, "lo": 0.0184, "hi": 0.2811, "ab": 0.0}, {"d": "2026-05-04", "acc": 0.1471, "n": 34, "lo": 0.0467, "hi": 0.3775, "ab": 0.0}, {"d": "2026-05-11", "acc": 0.1765, "n": 34, "lo": 0.0619, "hi": 0.4103, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.1154, "n": 26, "lo": 0.0271, "hi": 0.3791, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2026-06-01", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2026-06-08", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.0}, {"d": "2026-06-15", "acc": 0.0769, "n": 26, "lo": 0.0137, "hi": 0.3331, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.0385, "n": 26, "lo": 0.004, "hi": 0.2835, "ab": 0.0}, {"d": "2026-06-29", "acc": 0.1667, "n": 36, "lo": 0.0584, "hi": 0.3922, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.1471, "n": 34, "lo": 0.0467, "hi": 0.3775, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.2105, "n": 38, "lo": 0.0851, "hi": 0.4333, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.1429, "n": 28, "lo": 0.0401, "hi": 0.3994, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.25, "n": 16, "lo": 0.0715, "hi": 0.5907, "ab": 0.0}], "samples": [{"band": "mid", "date": "2024-08-03", "stem": "In early August 2024, Vietnam's ruling party formally installed a new General Secretary after the death of Nguyễn Phú Trọng the previous month. Who took the post?", "correct_opt": "Tô Lâm", "model_opt": "Tô Lâm", "raw": "C, 95", "ok": true, "abstain": false, "conf": 95}, {"band": "mid", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "the United Arab Emirates", "raw": "C, 85", "ok": false, "abstain": false, "conf": 85}, {"band": "mid", "date": "2024-08-05", "stem": "In early August 2024, separatist gunmen in Indonesia's Central Papua province shot dead a helicopter pilot moments after he touched down at a remote village, while the four Papuan passengers were freed unharmed. What wa…", "correct_opt": "New Zealander", "model_opt": "New Zealander", "raw": "G, 70", "ok": true, "abstain": false, "conf": 70}, {"band": "mid", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "X", "raw": "H, 85", "ok": false, "abstain": false, "conf": 85}, {"band": "post", "date": "2024-10-07", "stem": "In early October 2024, Russian missiles hit two cargo vessels in the Port of Odesa, killing a dock worker and wounding five crew members. Under which country's flag was the grain-carrying ship sailing?", "correct_opt": "Palau", "model_opt": "Saint Kitts and Nevis", "raw": "G, 60", "ok": false, "abstain": false, "conf": 60}, {"band": "post", "date": "2024-10-08", "stem": "In early October 2024, the Israeli military said an airstrike in Beirut had killed the man running Hezbollah's logistics headquarters, who also sat on the group's Jihad Council. Who was named as the slain official?", "correct_opt": "Suhail Hussein Husseini", "model_opt": "Ibrahim Aqil", "raw": "H, 73", "ok": false, "abstain": false, "conf": 73}, {"band": "post", "date": "2024-10-08", "stem": "In early October 2024, a Nordic country's security service raised its national terror alert from moderate to high, citing threats aimed at Jewish and Israeli sites. Which country raised its alert level?", "correct_opt": "Norway", "model_opt": "Sweden", "raw": "F, 85", "ok": false, "abstain": false, "conf": 85}]}, {"id": "gpt-5.1", "name": "GPT-5.1", "provider": "openai", "family": "gpt-5.1", "release": "2025-11-13", "pub_rel": "2024-09", "pub_train": "2024-09", "pub_rel_d": "2024-09-30", "pub_train_d": "2024-09-30", "fit": {"midpoint": "2024-09-05", "width": 120.0, "decay_start": "2023-11-19", "floor_reach": "2025-10-23", "floor": 0.1851, "plateau": 0.4828, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2024-09-05", "2024-09-05"]}, "weekly": [{"d": "2024-07-29", "acc": 0.8125, "n": 16, "lo": 0.4668, "hi": 0.9555, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.2692, "n": 26, "lo": 0.1036, "hi": 0.5402, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.25, "n": 24, "lo": 0.0889, "hi": 0.5323, "ab": 0.0}, {"d": "2024-08-19", "acc": 0.4615, "n": 26, "lo": 0.2321, "hi": 0.7086, "ab": 0.0}, {"d": "2024-08-26", "acc": 0.1818, "n": 22, "lo": 0.0514, "hi": 0.477, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.4375, "n": 32, "lo": 0.231, "hi": 0.6682, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.1538, "n": 26, "lo": 0.0433, "hi": 0.4224, "ab": 0.0}, {"d": "2024-09-16", "acc": 0.3214, "n": 28, "lo": 0.1398, "hi": 0.58, "ab": 0.0}, {"d": "2024-09-23", "acc": 0.5455, "n": 22, "lo": 0.2801, "hi": 0.7873, "ab": 0.0}, {"d": "2024-09-30", "acc": 0.25, "n": 24, "lo": 0.0889, "hi": 0.5323, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.1429, "n": 14, "lo": 0.0257, "hi": 0.5131, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.3636, "n": 22, "lo": 0.1517, "hi": 0.6462, "ab": 0.0}, {"d": "2024-10-21", "acc": 0.375, "n": 16, "lo": 0.1368, "hi": 0.6943, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.125, "n": 24, "lo": 0.0294, "hi": 0.4025, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.0}, {"d": "2024-11-11", "acc": 0.1667, "n": 24, "lo": 0.047, "hi": 0.448, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.1667, "n": 18, "lo": 0.0395, "hi": 0.4933, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.3529, "n": 34, "lo": 0.1731, "hi": 0.587, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.1818, "n": 22, "lo": 0.0514, "hi": 0.477, "ab": 0.0}, {"d": "2024-12-09", "acc": 0.0333, "n": 30, "lo": 0.0035, "hi": 0.2535, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.2105, "n": 38, "lo": 0.0851, "hi": 0.4333, "ab": 0.0}, {"d": "2024-12-23", "acc": 0.35, "n": 40, "lo": 0.1812, "hi": 0.5671, "ab": 0.0}, {"d": "2024-12-30", "acc": 0.3, "n": 10, "lo": 0.0726, "hi": 0.7012, "ab": 0.0}, {"d": "2025-01-06", "acc": 0.3333, "n": 18, "lo": 0.1206, "hi": 0.6458, "ab": 0.0}, {"d": "2025-01-13", "acc": 0.3125, "n": 16, "lo": 0.1024, "hi": 0.6442, "ab": 0.0}, {"d": "2025-01-20", "acc": 0.3636, "n": 22, "lo": 0.1517, "hi": 0.6462, "ab": 0.0}, {"d": "2025-01-27", "acc": 0.5333, "n": 30, "lo": 0.3012, "hi": 0.7519, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.1429, "n": 28, "lo": 0.0401, "hi": 0.3994, "ab": 0.0}, {"d": "2025-02-10", "acc": 0.0909, "n": 22, "lo": 0.0162, "hi": 0.3774, "ab": 0.0}, {"d": "2025-02-17", "acc": 0.2, "n": 30, "lo": 0.0705, "hi": 0.4519, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.5, "n": 28, "lo": 0.268, "hi": 0.732, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.2188, "n": 32, "lo": 0.0833, "hi": 0.4631, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.1667, "n": 36, "lo": 0.0584, "hi": 0.3922, "ab": 0.0}, {"d": "2025-03-17", "acc": 0.1154, "n": 26, "lo": 0.0271, "hi": 0.3791, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.2, "n": 20, "lo": 0.0567, "hi": 0.5098, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.0}, {"d": "2025-04-07", "acc": 0.2778, "n": 36, "lo": 0.125, "hi": 0.5087, "ab": 0.0}, {"d": "2025-04-14", "acc": 0.0455, "n": 22, "lo": 0.0048, "hi": 0.3215, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.2059, "n": 34, "lo": 0.0782, "hi": 0.442, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.0}, {"d": "2025-05-05", "acc": 0.2857, "n": 28, "lo": 0.1172, "hi": 0.5465, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.3158, "n": 38, "lo": 0.1536, "hi": 0.5399, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.3, "n": 30, "lo": 0.1299, "hi": 0.5517, "ab": 0.0}, {"d": "2025-05-26", "acc": 0.2917, "n": 24, "lo": 0.1127, "hi": 0.5716, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.3235, "n": 34, "lo": 0.1526, "hi": 0.5595, "ab": 0.0}, {"d": "2025-06-09", "acc": 0.25, "n": 24, "lo": 0.0889, "hi": 0.5323, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.3333, "n": 30, "lo": 0.1518, "hi": 0.5829, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.1471, "n": 34, "lo": 0.0467, "hi": 0.3775, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.1842, "n": 38, "lo": 0.0697, "hi": 0.4049, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.0833, "n": 24, "lo": 0.0149, "hi": 0.3539, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.2273, "n": 22, "lo": 0.0733, "hi": 0.5225, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.2667, "n": 30, "lo": 0.109, "hi": 0.5195, "ab": 0.0}, {"d": "2025-07-28", "acc": 0.1333, "n": 30, "lo": 0.0374, "hi": 0.3788, "ab": 0.0}, {"d": "2025-08-04", "acc": 0.0, "n": 18, "lo": 0.0, "hi": 0.2992, "ab": 0.0}, {"d": "2025-08-11", "acc": 0.2188, "n": 32, "lo": 0.0833, "hi": 0.4631, "ab": 0.0}, {"d": "2025-08-18", "acc": 0.3158, "n": 38, "lo": 0.1536, "hi": 0.5399, "ab": 0.0}, {"d": "2025-08-25", "acc": 0.1667, "n": 42, "lo": 0.0629, "hi": 0.3736, "ab": 0.0}, {"d": "2025-09-01", "acc": 0.35, "n": 20, "lo": 0.1369, "hi": 0.6463, "ab": 0.0}, {"d": "2025-09-08", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.0}, {"d": "2025-09-15", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.0}, {"d": "2025-09-22", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.0}, {"d": "2025-09-29", "acc": 0.0455, "n": 22, "lo": 0.0048, "hi": 0.3215, "ab": 0.0}, {"d": "2025-10-06", "acc": 0.2647, "n": 34, "lo": 0.1138, "hi": 0.5024, "ab": 0.0}, {"d": "2025-10-13", "acc": 0.0833, "n": 36, "lo": 0.0195, "hi": 0.2938, "ab": 0.0}, {"d": "2025-10-20", "acc": 0.225, "n": 40, "lo": 0.0959, "hi": 0.4427, "ab": 0.0}, {"d": "2025-10-27", "acc": 0.2143, "n": 42, "lo": 0.0912, "hi": 0.4258, "ab": 0.0}, {"d": "2025-11-03", "acc": 0.3462, "n": 26, "lo": 0.1513, "hi": 0.6112, "ab": 0.0}, {"d": "2025-11-10", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2025-11-17", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.2667, "n": 30, "lo": 0.109, "hi": 0.5195, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.2, "n": 30, "lo": 0.0705, "hi": 0.4519, "ab": 0.0}, {"d": "2025-12-08", "acc": 0.0909, "n": 22, "lo": 0.0162, "hi": 0.3774, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.2105, "n": 38, "lo": 0.0851, "hi": 0.4333, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.0455, "n": 22, "lo": 0.0048, "hi": 0.3215, "ab": 0.0}, {"d": "2025-12-29", "acc": 0.2917, "n": 24, "lo": 0.1127, "hi": 0.5716, "ab": 0.0}, {"d": "2026-01-05", "acc": 0.1154, "n": 26, "lo": 0.0271, "hi": 0.3791, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.2188, "n": 32, "lo": 0.0833, "hi": 0.4631, "ab": 0.0}, {"d": "2026-01-19", "acc": 0.0882, "n": 34, "lo": 0.0206, "hi": 0.3076, "ab": 0.0}, {"d": "2026-01-26", "acc": 0.1538, "n": 26, "lo": 0.0433, "hi": 0.4224, "ab": 0.0}, {"d": "2026-02-02", "acc": 0.0625, "n": 32, "lo": 0.0111, "hi": 0.2833, "ab": 0.0}, {"d": "2026-02-09", "acc": 0.1667, "n": 30, "lo": 0.0531, "hi": 0.4161, "ab": 0.0}, {"d": "2026-02-16", "acc": 0.0556, "n": 18, "lo": 0.0058, "hi": 0.3712, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.3235, "n": 34, "lo": 0.1526, "hi": 0.5595, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.1667, "n": 42, "lo": 0.0629, "hi": 0.3736, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.1429, "n": 28, "lo": 0.0401, "hi": 0.3994, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.3, "n": 20, "lo": 0.1078, "hi": 0.6032, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.2812, "n": 32, "lo": 0.1213, "hi": 0.5259, "ab": 0.0}, {"d": "2026-04-06", "acc": 0.2941, "n": 34, "lo": 0.1328, "hi": 0.5313, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.2647, "n": 34, "lo": 0.1138, "hi": 0.5024, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.1, "n": 30, "lo": 0.0234, "hi": 0.3397, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.0526, "n": 38, "lo": 0.0094, "hi": 0.2464, "ab": 0.0}, {"d": "2026-05-04", "acc": 0.3235, "n": 34, "lo": 0.1526, "hi": 0.5595, "ab": 0.0}, {"d": "2026-05-11", "acc": 0.1176, "n": 34, "lo": 0.0329, "hi": 0.3434, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.1538, "n": 26, "lo": 0.0433, "hi": 0.4224, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.0}, {"d": "2026-06-01", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.0}, {"d": "2026-06-08", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.0}, {"d": "2026-06-15", "acc": 0.1538, "n": 26, "lo": 0.0433, "hi": 0.4224, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.0769, "n": 26, "lo": 0.0137, "hi": 0.3331, "ab": 0.0}, {"d": "2026-06-29", "acc": 0.1667, "n": 36, "lo": 0.0584, "hi": 0.3922, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.2059, "n": 34, "lo": 0.0782, "hi": 0.442, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.1316, "n": 38, "lo": 0.0417, "hi": 0.3454, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.0625, "n": 16, "lo": 0.0066, "hi": 0.4023, "ab": 0.0}], "samples": [{"band": "mid", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 92", "ok": true, "abstain": false, "conf": 92}, {"band": "mid", "date": "2024-08-02", "stem": "In early August 2024, a South American government became the third country in the world to recognise Edmundo González as the rightful winner of Venezuela's disputed presidential election. Which country was it?", "correct_opt": "Argentina", "model_opt": "Uruguay", "raw": "C, 55", "ok": false, "abstain": false, "conf": 55}, {"band": "mid", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 60", "ok": true, "abstain": false, "conf": 60}, {"band": "mid", "date": "2024-08-01", "stem": "In early August 2024, an Eastern European government expelled a Russian diplomat, saying the envoy had been engaged in spying. Which country ordered the expulsion?", "correct_opt": "Moldova", "model_opt": "Latvia", "raw": "C, 60", "ok": false, "abstain": false, "conf": 60}, {"band": "post", "date": "2025-10-23", "stem": "In late October 2025, U.S. authorities arrested more than thirty people in a mafia-linked illegal gambling case that ensnared figures from professional basketball, including a serving head coach. Which team's head coach…", "correct_opt": "the Portland Trail Blazers", "model_opt": "the Cleveland Cavaliers", "raw": "B, 20", "ok": false, "abstain": false, "conf": 20}, {"band": "post", "date": "2025-10-23", "stem": "In late October 2025, Brazil's president publicly confirmed he would seek another term in the following year's election, doing so during a meeting with a visiting head of state. Which leader was he meeting?", "correct_opt": "Prabowo Subianto of Indonesia", "model_opt": "Gustavo Petro of Colombia", "raw": "C, 35", "ok": false, "abstain": false, "conf": 35}, {"band": "post", "date": "2025-10-24", "stem": "In late October 2025, a roadside bomb struck a police vehicle in a district of Khyber Pakhtunkhwa that had once been a bastion of the Pakistani Taliban, killing three people. In which district did the blast occur?", "correct_opt": "Hangu", "model_opt": "Lakki Marwat", "raw": "C, 40", "ok": false, "abstain": false, "conf": 40}]}, {"id": "gpt-5.2", "name": "GPT-5.2", "provider": "openai", "family": "gpt-5.2", "release": "2025-12-11", "pub_rel": "2025-08", "pub_train": "2025-08", "pub_rel_d": "2025-08-31", "pub_train_d": "2025-08-31", "fit": {"midpoint": "2025-08-12", "width": 33.96, "decay_start": "2025-05-18", "floor_reach": "2025-12-16", "floor": 0.2603, "plateau": 0.809, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2025-07-26", "2025-08-31"]}, "weekly": [{"d": "2024-07-29", "acc": 0.8125, "n": 16, "lo": 0.4668, "hi": 0.9555, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.7692, "n": 26, "lo": 0.4974, "hi": 0.9182, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.625, "n": 24, "lo": 0.3544, "hi": 0.835, "ab": 0.0}, {"d": "2024-08-19", "acc": 0.8077, "n": 26, "lo": 0.5366, "hi": 0.9384, "ab": 0.0}, {"d": "2024-08-26", "acc": 0.7273, "n": 22, "lo": 0.4343, "hi": 0.9025, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.8125, "n": 32, "lo": 0.5699, "hi": 0.9341, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.6923, "n": 26, "lo": 0.4237, "hi": 0.8732, "ab": 0.0}, {"d": "2024-09-16", "acc": 0.75, "n": 28, "lo": 0.4882, "hi": 0.9042, "ab": 0.0}, {"d": "2024-09-23", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2024-09-30", "acc": 0.8333, "n": 24, "lo": 0.552, "hi": 0.953, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.7143, "n": 14, "lo": 0.3589, "hi": 0.9178, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.6818, "n": 22, "lo": 0.3932, "hi": 0.8764, "ab": 0.0}, {"d": "2024-10-21", "acc": 0.6875, "n": 16, "lo": 0.3558, "hi": 0.8976, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.625, "n": 24, "lo": 0.3544, "hi": 0.835, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.7273, "n": 22, "lo": 0.4343, "hi": 0.9025, "ab": 0.0}, {"d": "2024-11-11", "acc": 0.875, "n": 24, "lo": 0.5975, "hi": 0.9706, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.8333, "n": 18, "lo": 0.5067, "hi": 0.9605, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.7353, "n": 34, "lo": 0.4976, "hi": 0.8862, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.5455, "n": 22, "lo": 0.2801, "hi": 0.7873, "ab": 0.0}, {"d": "2024-12-09", "acc": 0.8667, "n": 30, "lo": 0.6212, "hi": 0.9626, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.7632, "n": 38, "lo": 0.539, "hi": 0.8988, "ab": 0.0}, {"d": "2024-12-23", "acc": 0.875, "n": 40, "lo": 0.6687, "hi": 0.9604, "ab": 0.0}, {"d": "2024-12-30", "acc": 1.0, "n": 10, "lo": 0.5655, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-06", "acc": 1.0, "n": 18, "lo": 0.7008, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-13", "acc": 0.8125, "n": 16, "lo": 0.4668, "hi": 0.9555, "ab": 0.0}, {"d": "2025-01-20", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-27", "acc": 0.9, "n": 30, "lo": 0.6603, "hi": 0.9766, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.8571, "n": 28, "lo": 0.6006, "hi": 0.9599, "ab": 0.0}, {"d": "2025-02-10", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2025-02-17", "acc": 0.8, "n": 30, "lo": 0.5481, "hi": 0.9295, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.8571, "n": 28, "lo": 0.6006, "hi": 0.9599, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.9062, "n": 32, "lo": 0.6771, "hi": 0.978, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.7222, "n": 36, "lo": 0.4913, "hi": 0.875, "ab": 0.0}, {"d": "2025-03-17", "acc": 0.8846, "n": 26, "lo": 0.6209, "hi": 0.9729, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.75, "n": 20, "lo": 0.4422, "hi": 0.9191, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.8929, "n": 28, "lo": 0.6417, "hi": 0.9749, "ab": 0.0}, {"d": "2025-04-07", "acc": 0.8333, "n": 36, "lo": 0.6078, "hi": 0.9416, "ab": 0.0}, {"d": "2025-04-14", "acc": 0.7727, "n": 22, "lo": 0.4775, "hi": 0.9267, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.8824, "n": 34, "lo": 0.6566, "hi": 0.9671, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.75, "n": 28, "lo": 0.4882, "hi": 0.9042, "ab": 0.0}, {"d": "2025-05-05", "acc": 0.7857, "n": 28, "lo": 0.5241, "hi": 0.9243, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.8158, "n": 38, "lo": 0.5951, "hi": 0.9303, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.8667, "n": 30, "lo": 0.6212, "hi": 0.9626, "ab": 0.0}, {"d": "2025-05-26", "acc": 0.8333, "n": 24, "lo": 0.552, "hi": 0.953, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.6471, "n": 34, "lo": 0.413, "hi": 0.8269, "ab": 0.0}, {"d": "2025-06-09", "acc": 0.75, "n": 24, "lo": 0.4677, "hi": 0.9111, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.6667, "n": 30, "lo": 0.4171, "hi": 0.8482, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.7353, "n": 34, "lo": 0.4976, "hi": 0.8862, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.5789, "n": 38, "lo": 0.3628, "hi": 0.7686, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.5833, "n": 24, "lo": 0.3195, "hi": 0.8067, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.7727, "n": 22, "lo": 0.4775, "hi": 0.9267, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.6667, "n": 30, "lo": 0.4171, "hi": 0.8482, "ab": 0.0}, {"d": "2025-07-28", "acc": 0.7667, "n": 30, "lo": 0.5137, "hi": 0.9109, "ab": 0.0}, {"d": "2025-08-04", "acc": 0.5, "n": 18, "lo": 0.2265, "hi": 0.7735, "ab": 0.0}, {"d": "2025-08-11", "acc": 0.5, "n": 32, "lo": 0.28, "hi": 0.72, "ab": 0.0}, {"d": "2025-08-18", "acc": 0.4211, "n": 38, "lo": 0.2314, "hi": 0.6372, "ab": 0.0}, {"d": "2025-08-25", "acc": 0.3095, "n": 42, "lo": 0.1548, "hi": 0.5231, "ab": 0.0}, {"d": "2025-09-01", "acc": 0.5, "n": 20, "lo": 0.2366, "hi": 0.7634, "ab": 0.0}, {"d": "2025-09-08", "acc": 0.4286, "n": 28, "lo": 0.2138, "hi": 0.6741, "ab": 0.0}, {"d": "2025-09-15", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.0}, {"d": "2025-09-22", "acc": 0.3636, "n": 22, "lo": 0.1517, "hi": 0.6462, "ab": 0.0}, {"d": "2025-09-29", "acc": 0.2273, "n": 22, "lo": 0.0733, "hi": 0.5225, "ab": 0.0}, {"d": "2025-10-06", "acc": 0.5, "n": 34, "lo": 0.2853, "hi": 0.7147, "ab": 0.0}, {"d": "2025-10-13", "acc": 0.3889, "n": 36, "lo": 0.203, "hi": 0.6138, "ab": 0.0}, {"d": "2025-10-20", "acc": 0.25, "n": 40, "lo": 0.1119, "hi": 0.4687, "ab": 0.0}, {"d": "2025-10-27", "acc": 0.4762, "n": 42, "lo": 0.2834, "hi": 0.6763, "ab": 0.0}, {"d": "2025-11-03", "acc": 0.5769, "n": 26, "lo": 0.3228, "hi": 0.796, "ab": 0.0}, {"d": "2025-11-10", "acc": 0.3214, "n": 28, "lo": 0.1398, "hi": 0.58, "ab": 0.0}, {"d": "2025-11-17", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.2333, "n": 30, "lo": 0.0891, "hi": 0.4863, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.3667, "n": 30, "lo": 0.1746, "hi": 0.6131, "ab": 0.0}, {"d": "2025-12-08", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.2895, "n": 38, "lo": 0.1356, "hi": 0.5142, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.0}, {"d": "2025-12-29", "acc": 0.1667, "n": 24, "lo": 0.047, "hi": 0.448, "ab": 0.0}, {"d": "2026-01-05", "acc": 0.2308, "n": 26, "lo": 0.0818, "hi": 0.5026, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.3125, "n": 32, "lo": 0.1416, "hi": 0.556, "ab": 0.0}, {"d": "2026-01-19", "acc": 0.2059, "n": 34, "lo": 0.0782, "hi": 0.442, "ab": 0.0}, {"d": "2026-01-26", "acc": 0.1538, "n": 26, "lo": 0.0433, "hi": 0.4224, "ab": 0.0}, {"d": "2026-02-02", "acc": 0.2188, "n": 32, "lo": 0.0833, "hi": 0.4631, "ab": 0.0}, {"d": "2026-02-09", "acc": 0.3, "n": 30, "lo": 0.1299, "hi": 0.5517, "ab": 0.0}, {"d": "2026-02-16", "acc": 0.2778, "n": 18, "lo": 0.0904, "hi": 0.5981, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.3529, "n": 34, "lo": 0.1731, "hi": 0.587, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.3095, "n": 42, "lo": 0.1548, "hi": 0.5231, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.45, "n": 20, "lo": 0.2014, "hi": 0.7263, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.4375, "n": 32, "lo": 0.231, "hi": 0.6682, "ab": 0.0}, {"d": "2026-04-06", "acc": 0.3529, "n": 34, "lo": 0.1731, "hi": 0.587, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.2941, "n": 34, "lo": 0.1328, "hi": 0.5313, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.2, "n": 30, "lo": 0.0705, "hi": 0.4519, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.1316, "n": 38, "lo": 0.0417, "hi": 0.3454, "ab": 0.0}, {"d": "2026-05-04", "acc": 0.5, "n": 34, "lo": 0.2853, "hi": 0.7147, "ab": 0.0}, {"d": "2026-05-11", "acc": 0.1176, "n": 34, "lo": 0.0329, "hi": 0.3434, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.2308, "n": 26, "lo": 0.0818, "hi": 0.5026, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.0714, "n": 28, "lo": 0.0127, "hi": 0.3147, "ab": 0.0}, {"d": "2026-06-01", "acc": 0.2857, "n": 28, "lo": 0.1172, "hi": 0.5465, "ab": 0.0}, {"d": "2026-06-08", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.0}, {"d": "2026-06-15", "acc": 0.2692, "n": 26, "lo": 0.1036, "hi": 0.5402, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.1923, "n": 26, "lo": 0.0616, "hi": 0.4634, "ab": 0.0}, {"d": "2026-06-29", "acc": 0.3889, "n": 36, "lo": 0.203, "hi": 0.6138, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.0588, "n": 34, "lo": 0.0105, "hi": 0.2698, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.1316, "n": 38, "lo": 0.0417, "hi": 0.3454, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.1875, "n": 16, "lo": 0.0445, "hi": 0.5332, "ab": 0.0}], "samples": [{"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 92", "ok": true, "abstain": false, "conf": 92}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 95", "ok": true, "abstain": false, "conf": 95}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, a South American government became the third country in the world to recognise Edmundo González as the rightful winner of Venezuela's disputed presidential election. Which country was it?", "correct_opt": "Argentina", "model_opt": "Paraguay", "raw": "B, 55", "ok": false, "abstain": false, "conf": 55}, {"band": "mid", "date": "2025-05-20", "stem": "In mid-May 2025, an Indian writer and her translator shared the International Booker Prize for a collection of short stories. What was the title of the winning book?", "correct_opt": "Heart Lamp", "model_opt": "Heart Lamp", "raw": "E, 92", "ok": true, "abstain": false, "conf": 92}, {"band": "mid", "date": "2025-05-24", "stem": "In late May 2025, a Cypriot Cup final at the GSP Stadium in Nicosia finished goalless and was settled 5–4 on penalties. Which club lifted the trophy?", "correct_opt": "AEK Larnaca", "model_opt": "Pafos FC", "raw": "G, 55", "ok": false, "abstain": false, "conf": 55}, {"band": "mid", "date": "2025-05-20", "stem": "In mid-May 2025, three pupils were wounded in a knife attack at a school by a 16-year-old who had published a manifesto declaring women his intended targets. In which Finnish municipality did the attack take place?", "correct_opt": "Pirkkala", "model_opt": "Pirkkala", "raw": "B, 78", "ok": true, "abstain": false, "conf": 78}, {"band": "mid", "date": "2025-05-24", "stem": "In late May 2025, the German Cup final at Berlin's Olympiastadion ended 4–2 in favour of a Bundesliga side over third-tier opponents. Which club were the beaten finalists?", "correct_opt": "Arminia Bielefeld", "model_opt": "Saarbrücken", "raw": "B, 35", "ok": false, "abstain": false, "conf": 35}, {"band": "post", "date": "2025-12-16", "stem": "In mid-December 2025, the March 23 Movement said it would pull its fighters out of a town it had recently taken in South Kivu, saying it was acting on a request from Washington. Which town was to be vacated?", "correct_opt": "Uvira", "model_opt": "Minova", "raw": "A, 62", "ok": false, "abstain": false, "conf": 62}, {"band": "post", "date": "2025-12-16", "stem": "In mid-December 2025, Honduran president Xiomara Castro said she would take her complaint of an alleged \"electoral coup\" to international bodies, and accused a foreign leader of meddling in the vote. Whom did she name?", "correct_opt": "Donald Trump", "model_opt": "Nayib Bukele", "raw": "E, 45", "ok": false, "abstain": false, "conf": 45}, {"band": "post", "date": "2025-12-16", "stem": "In mid-December 2025, FIFA named its men's player of the year for 2025, choosing a French forward. Which player received the award?", "correct_opt": "Ousmane Dembélé", "model_opt": "Kylian Mbappé", "raw": "A, 62", "ok": false, "abstain": false, "conf": 62}]}, {"id": "gpt-5.3-codex", "name": "GPT-5.3-Codex", "provider": "openai", "family": "gpt-5.3", "release": "2026-02-05", "pub_rel": "2025-08", "pub_train": "2025-08", "pub_rel_d": "2025-08-31", "pub_train_d": "2025-08-31", "fit": {"midpoint": "2025-08-13", "width": 18.64, "decay_start": "2025-06-28", "floor_reach": "2025-10-19", "floor": 0.3239, "plateau": 0.9237, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2025-08-05", "2025-08-27"]}, "weekly": [{"d": "2024-07-29", "acc": 0.8125, "n": 16, "lo": 0.4668, "hi": 0.9555, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.8462, "n": 26, "lo": 0.5776, "hi": 0.9567, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.9167, "n": 24, "lo": 0.6461, "hi": 0.9851, "ab": 0.0}, {"d": "2024-08-19", "acc": 0.9231, "n": 26, "lo": 0.6669, "hi": 0.9863, "ab": 0.0}, {"d": "2024-08-26", "acc": 0.7727, "n": 22, "lo": 0.4775, "hi": 0.9267, "ab": 0.0}, {"d": "2024-09-02", "acc": 1.0, "n": 32, "lo": 0.8064, "hi": 1.0, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.7692, "n": 26, "lo": 0.4974, "hi": 0.9182, "ab": 0.0}, {"d": "2024-09-16", "acc": 0.8571, "n": 28, "lo": 0.6006, "hi": 0.9599, "ab": 0.0}, {"d": "2024-09-23", "acc": 0.9545, "n": 22, "lo": 0.6785, "hi": 0.9952, "ab": 0.0}, {"d": "2024-09-30", "acc": 0.9167, "n": 24, "lo": 0.6461, "hi": 0.9851, "ab": 0.0}, {"d": "2024-10-07", "acc": 1.0, "n": 14, "lo": 0.6457, "hi": 1.0, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.7727, "n": 22, "lo": 0.4775, "hi": 0.9267, "ab": 0.0}, {"d": "2024-10-21", "acc": 0.9375, "n": 16, "lo": 0.5977, "hi": 0.9934, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.875, "n": 24, "lo": 0.5975, "hi": 0.9706, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2024-11-11", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.8889, "n": 18, "lo": 0.565, "hi": 0.9801, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.8529, "n": 34, "lo": 0.6225, "hi": 0.9533, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2024-12-09", "acc": 0.9333, "n": 30, "lo": 0.7018, "hi": 0.9881, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.9211, "n": 38, "lo": 0.7189, "hi": 0.9816, "ab": 0.0}, {"d": "2024-12-23", "acc": 1.0, "n": 40, "lo": 0.8389, "hi": 1.0, "ab": 0.0}, {"d": "2024-12-30", "acc": 1.0, "n": 10, "lo": 0.5655, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-06", "acc": 1.0, "n": 18, "lo": 0.7008, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-13", "acc": 1.0, "n": 16, "lo": 0.6756, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-20", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-27", "acc": 1.0, "n": 30, "lo": 0.7961, "hi": 1.0, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.8929, "n": 28, "lo": 0.6417, "hi": 0.9749, "ab": 0.0}, {"d": "2025-02-10", "acc": 0.9545, "n": 22, "lo": 0.6785, "hi": 0.9952, "ab": 0.0}, {"d": "2025-02-17", "acc": 0.9333, "n": 30, "lo": 0.7018, "hi": 0.9881, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.9643, "n": 28, "lo": 0.7324, "hi": 0.9963, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.9688, "n": 32, "lo": 0.7593, "hi": 0.9967, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.9444, "n": 36, "lo": 0.7424, "hi": 0.9901, "ab": 0.0}, {"d": "2025-03-17", "acc": 1.0, "n": 26, "lo": 0.7719, "hi": 1.0, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.9, "n": 20, "lo": 0.5958, "hi": 0.9821, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-04-07", "acc": 0.9722, "n": 36, "lo": 0.7812, "hi": 0.9971, "ab": 0.0}, {"d": "2025-04-14", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.9118, "n": 34, "lo": 0.6924, "hi": 0.9794, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-05-05", "acc": 0.8929, "n": 28, "lo": 0.6417, "hi": 0.9749, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.8947, "n": 38, "lo": 0.6861, "hi": 0.9706, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.9667, "n": 30, "lo": 0.7465, "hi": 0.9965, "ab": 0.0}, {"d": "2025-05-26", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.8529, "n": 34, "lo": 0.6225, "hi": 0.9533, "ab": 0.0}, {"d": "2025-06-09", "acc": 0.9167, "n": 24, "lo": 0.6461, "hi": 0.9851, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.7667, "n": 30, "lo": 0.5137, "hi": 0.9109, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.9706, "n": 34, "lo": 0.7708, "hi": 0.9969, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.8684, "n": 38, "lo": 0.6546, "hi": 0.9583, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.875, "n": 24, "lo": 0.5975, "hi": 0.9706, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.8, "n": 30, "lo": 0.5481, "hi": 0.9295, "ab": 0.0}, {"d": "2025-07-28", "acc": 0.7333, "n": 30, "lo": 0.4805, "hi": 0.891, "ab": 0.0}, {"d": "2025-08-04", "acc": 0.8333, "n": 18, "lo": 0.5067, "hi": 0.9605, "ab": 0.0}, {"d": "2025-08-11", "acc": 0.4688, "n": 32, "lo": 0.2551, "hi": 0.6945, "ab": 0.0}, {"d": "2025-08-18", "acc": 0.4737, "n": 38, "lo": 0.2733, "hi": 0.6829, "ab": 0.0}, {"d": "2025-08-25", "acc": 0.4524, "n": 42, "lo": 0.2639, "hi": 0.6556, "ab": 0.0}, {"d": "2025-09-01", "acc": 0.5, "n": 20, "lo": 0.2366, "hi": 0.7634, "ab": 0.0}, {"d": "2025-09-08", "acc": 0.5357, "n": 28, "lo": 0.2965, "hi": 0.7596, "ab": 0.0}, {"d": "2025-09-15", "acc": 0.4091, "n": 22, "lo": 0.1814, "hi": 0.6839, "ab": 0.0}, {"d": "2025-09-22", "acc": 0.4545, "n": 22, "lo": 0.2127, "hi": 0.7199, "ab": 0.0}, {"d": "2025-09-29", "acc": 0.0909, "n": 22, "lo": 0.0162, "hi": 0.3774, "ab": 0.0}, {"d": "2025-10-06", "acc": 0.5, "n": 34, "lo": 0.2853, "hi": 0.7147, "ab": 0.0}, {"d": "2025-10-13", "acc": 0.5, "n": 36, "lo": 0.2903, "hi": 0.7097, "ab": 0.0}, {"d": "2025-10-20", "acc": 0.35, "n": 40, "lo": 0.1812, "hi": 0.5671, "ab": 0.0}, {"d": "2025-10-27", "acc": 0.5238, "n": 42, "lo": 0.3237, "hi": 0.7166, "ab": 0.0}, {"d": "2025-11-03", "acc": 0.5769, "n": 26, "lo": 0.3228, "hi": 0.796, "ab": 0.0}, {"d": "2025-11-10", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.0}, {"d": "2025-11-17", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.3333, "n": 30, "lo": 0.1518, "hi": 0.5829, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.4333, "n": 30, "lo": 0.2228, "hi": 0.6711, "ab": 0.0}, {"d": "2025-12-08", "acc": 0.4545, "n": 22, "lo": 0.2127, "hi": 0.7199, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.2368, "n": 38, "lo": 0.1012, "hi": 0.461, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.2273, "n": 22, "lo": 0.0733, "hi": 0.5225, "ab": 0.0}, {"d": "2025-12-29", "acc": 0.3333, "n": 24, "lo": 0.1381, "hi": 0.6094, "ab": 0.0}, {"d": "2026-01-05", "acc": 0.1923, "n": 26, "lo": 0.0616, "hi": 0.4634, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.375, "n": 32, "lo": 0.1848, "hi": 0.6136, "ab": 0.0}, {"d": "2026-01-19", "acc": 0.3824, "n": 34, "lo": 0.1943, "hi": 0.6138, "ab": 0.0}, {"d": "2026-01-26", "acc": 0.1923, "n": 26, "lo": 0.0616, "hi": 0.4634, "ab": 0.0}, {"d": "2026-02-02", "acc": 0.2812, "n": 32, "lo": 0.1213, "hi": 0.5259, "ab": 0.0}, {"d": "2026-02-09", "acc": 0.3667, "n": 30, "lo": 0.1746, "hi": 0.6131, "ab": 0.0}, {"d": "2026-02-16", "acc": 0.1667, "n": 18, "lo": 0.0395, "hi": 0.4933, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.4706, "n": 34, "lo": 0.2616, "hi": 0.6904, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.2857, "n": 42, "lo": 0.1381, "hi": 0.4996, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.45, "n": 20, "lo": 0.2014, "hi": 0.7263, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.4062, "n": 32, "lo": 0.2075, "hi": 0.6413, "ab": 0.0}, {"d": "2026-04-06", "acc": 0.3824, "n": 34, "lo": 0.1943, "hi": 0.6138, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.3824, "n": 34, "lo": 0.1943, "hi": 0.6138, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.2, "n": 30, "lo": 0.0705, "hi": 0.4519, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.2368, "n": 38, "lo": 0.1012, "hi": 0.461, "ab": 0.0}, {"d": "2026-05-04", "acc": 0.5294, "n": 34, "lo": 0.3096, "hi": 0.7384, "ab": 0.0}, {"d": "2026-05-11", "acc": 0.3235, "n": 34, "lo": 0.1526, "hi": 0.5595, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.3462, "n": 26, "lo": 0.1513, "hi": 0.6112, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.0714, "n": 28, "lo": 0.0127, "hi": 0.3147, "ab": 0.0}, {"d": "2026-06-01", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.0}, {"d": "2026-06-08", "acc": 0.3636, "n": 22, "lo": 0.1517, "hi": 0.6462, "ab": 0.0}, {"d": "2026-06-15", "acc": 0.3846, "n": 26, "lo": 0.1771, "hi": 0.6448, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.1538, "n": 26, "lo": 0.0433, "hi": 0.4224, "ab": 0.0}, {"d": "2026-06-29", "acc": 0.4444, "n": 36, "lo": 0.2456, "hi": 0.6628, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.0882, "n": 34, "lo": 0.0206, "hi": 0.3076, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.2105, "n": 38, "lo": 0.0851, "hi": 0.4333, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.4286, "n": 28, "lo": 0.2138, "hi": 0.6741, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.375, "n": 16, "lo": 0.1368, "hi": 0.6943, "ab": 0.0}], "samples": [{"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 98", "ok": true, "abstain": false, "conf": 98}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 98", "ok": true, "abstain": false, "conf": 98}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, a South American government became the third country in the world to recognise Edmundo González as the rightful winner of Venezuela's disputed presidential election. Which country was it?", "correct_opt": "Argentina", "model_opt": "Ecuador", "raw": "F, 74", "ok": false, "abstain": false, "conf": 74}, {"band": "mid", "date": "2025-06-30", "stem": "In late June 2025, a 92-year-old man was found guilty of a 1967 killing described as the oldest cold case ever cracked in Britain. What was the name of the woman he was convicted of murdering?", "correct_opt": "Louisa Dunne", "model_opt": "Louisa Dunne", "raw": "D, 98", "ok": true, "abstain": false, "conf": 98}, {"band": "mid", "date": "2025-06-30", "stem": "In late June 2025, Washington agreed to resume stalled trade negotiations with Canada after Ottawa dropped a tax that had angered American technology companies. Which U.S. official publicly announced the restart of the …", "correct_opt": "National Economic Council director Kevin Hassett", "model_opt": "White House press secretary Karoline Leavitt", "raw": "B, 58", "ok": false, "abstain": false, "conf": 58}, {"band": "mid", "date": "2025-07-01", "stem": "In early July 2025, Beijing imposed an entry ban on a former Philippine Senate majority leader over his activities relating to China. Who was sanctioned?", "correct_opt": "Francis Tolentino", "model_opt": "Francis Tolentino", "raw": "E, 86", "ok": true, "abstain": false, "conf": 86}, {"band": "mid", "date": "2025-07-01", "stem": "In early July 2025, an overnight fire damaged the oldest line of the Warsaw Metro. What did the city's mayor identify as the cause of the incident?", "correct_opt": "A power outage", "model_opt": "An arson attack", "raw": "C, 82", "ok": false, "abstain": false, "conf": 82}, {"band": "post", "date": "2025-10-20", "stem": "In late October 2025, a cargo jet overran a runway at Hong Kong International Airport, struck a ground vehicle killing two workers, and ended up in the water. Which airline operated the freighter flight?", "correct_opt": "Emirates SkyCargo", "model_opt": "Cargolux", "raw": "B, 23", "ok": false, "abstain": false, "conf": 23}, {"band": "post", "date": "2025-10-20", "stem": "In late October 2025, an unusual tornado struck a commune north of Paris in the Val-d'Oise department, killing one person and leaving four others in critical condition. Which commune was hit?", "correct_opt": "Ermont", "model_opt": "Persan", "raw": "A, 42", "ok": false, "abstain": false, "conf": 42}, {"band": "post", "date": "2025-10-20", "stem": "In late October 2025, Lithuania briefly halted all flights at Vilnius airport and shut its airspace and border crossings. What prompted the government's decision?", "correct_opt": "sightings of smuggling balloons drifting in from Belarus", "model_opt": "reports of a Russian drone crossing from Kaliningrad", "raw": "A, 82", "ok": false, "abstain": false, "conf": 82}]}, {"id": "gpt-5.4", "name": "GPT-5.4", "provider": "openai", "family": "gpt-5.4", "release": "2026-03-05", "pub_rel": "2025-08", "pub_train": "2025-08", "pub_rel_d": "2025-08-31", "pub_train_d": "2025-08-31", "fit": {"midpoint": "2025-08-21", "width": 27.06, "decay_start": "2025-06-16", "floor_reach": "2025-11-25", "floor": 0.3385, "plateau": 0.9336, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2025-08-09", "2025-09-04"]}, "weekly": [{"d": "2024-07-29", "acc": 0.8125, "n": 16, "lo": 0.4668, "hi": 0.9555, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.9615, "n": 26, "lo": 0.7165, "hi": 0.996, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.9167, "n": 24, "lo": 0.6461, "hi": 0.9851, "ab": 0.0}, {"d": "2024-08-19", "acc": 0.9231, "n": 26, "lo": 0.6669, "hi": 0.9863, "ab": 0.0}, {"d": "2024-08-26", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.9688, "n": 32, "lo": 0.7593, "hi": 0.9967, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.7308, "n": 26, "lo": 0.4598, "hi": 0.8964, "ab": 0.0}, {"d": "2024-09-16", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2024-09-23", "acc": 0.9545, "n": 22, "lo": 0.6785, "hi": 0.9952, "ab": 0.0}, {"d": "2024-09-30", "acc": 0.9583, "n": 24, "lo": 0.6987, "hi": 0.9956, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.9286, "n": 14, "lo": 0.5609, "hi": 0.9925, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2024-10-21", "acc": 1.0, "n": 16, "lo": 0.6756, "hi": 1.0, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.8333, "n": 24, "lo": 0.552, "hi": 0.953, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2024-11-11", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.8889, "n": 18, "lo": 0.565, "hi": 0.9801, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.8235, "n": 34, "lo": 0.5897, "hi": 0.9381, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2024-12-09", "acc": 0.9667, "n": 30, "lo": 0.7465, "hi": 0.9965, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.8947, "n": 38, "lo": 0.6861, "hi": 0.9706, "ab": 0.0}, {"d": "2024-12-23", "acc": 0.95, "n": 40, "lo": 0.7639, "hi": 0.9911, "ab": 0.0}, {"d": "2024-12-30", "acc": 1.0, "n": 10, "lo": 0.5655, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-06", "acc": 1.0, "n": 18, "lo": 0.7008, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-13", "acc": 0.875, "n": 16, "lo": 0.5291, "hi": 0.9776, "ab": 0.0}, {"d": "2025-01-20", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-27", "acc": 1.0, "n": 30, "lo": 0.7961, "hi": 1.0, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-02-10", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-02-17", "acc": 0.9667, "n": 30, "lo": 0.7465, "hi": 0.9965, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.9643, "n": 28, "lo": 0.7324, "hi": 0.9963, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.9375, "n": 32, "lo": 0.7167, "hi": 0.9889, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.9444, "n": 36, "lo": 0.7424, "hi": 0.9901, "ab": 0.0}, {"d": "2025-03-17", "acc": 1.0, "n": 26, "lo": 0.7719, "hi": 1.0, "ab": 0.0}, {"d": "2025-03-24", "acc": 1.0, "n": 20, "lo": 0.7225, "hi": 1.0, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.9643, "n": 28, "lo": 0.7324, "hi": 0.9963, "ab": 0.0}, {"d": "2025-04-07", "acc": 1.0, "n": 36, "lo": 0.8241, "hi": 1.0, "ab": 0.0}, {"d": "2025-04-14", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.9412, "n": 34, "lo": 0.7302, "hi": 0.9895, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-05-05", "acc": 0.9643, "n": 28, "lo": 0.7324, "hi": 0.9963, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.8421, "n": 38, "lo": 0.6243, "hi": 0.9448, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.9667, "n": 30, "lo": 0.7465, "hi": 0.9965, "ab": 0.0}, {"d": "2025-05-26", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.8529, "n": 34, "lo": 0.6225, "hi": 0.9533, "ab": 0.0}, {"d": "2025-06-09", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.7333, "n": 30, "lo": 0.4805, "hi": 0.891, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.8824, "n": 34, "lo": 0.6566, "hi": 0.9671, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.8947, "n": 38, "lo": 0.6861, "hi": 0.9706, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.8333, "n": 24, "lo": 0.552, "hi": 0.953, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.8667, "n": 30, "lo": 0.6212, "hi": 0.9626, "ab": 0.0}, {"d": "2025-07-28", "acc": 0.8, "n": 30, "lo": 0.5481, "hi": 0.9295, "ab": 0.0}, {"d": "2025-08-04", "acc": 0.7778, "n": 18, "lo": 0.4526, "hi": 0.9368, "ab": 0.0}, {"d": "2025-08-11", "acc": 0.5312, "n": 32, "lo": 0.3055, "hi": 0.7449, "ab": 0.0}, {"d": "2025-08-18", "acc": 0.5, "n": 38, "lo": 0.2949, "hi": 0.7051, "ab": 0.0}, {"d": "2025-08-25", "acc": 0.5, "n": 42, "lo": 0.3034, "hi": 0.6966, "ab": 0.0}, {"d": "2025-09-01", "acc": 0.55, "n": 20, "lo": 0.2737, "hi": 0.7986, "ab": 0.0}, {"d": "2025-09-08", "acc": 0.6786, "n": 28, "lo": 0.42, "hi": 0.8602, "ab": 0.0}, {"d": "2025-09-15", "acc": 0.4545, "n": 22, "lo": 0.2127, "hi": 0.7199, "ab": 0.0}, {"d": "2025-09-22", "acc": 0.5, "n": 22, "lo": 0.2456, "hi": 0.7544, "ab": 0.0}, {"d": "2025-09-29", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.0}, {"d": "2025-10-06", "acc": 0.5, "n": 34, "lo": 0.2853, "hi": 0.7147, "ab": 0.0}, {"d": "2025-10-13", "acc": 0.5, "n": 36, "lo": 0.2903, "hi": 0.7097, "ab": 0.0}, {"d": "2025-10-20", "acc": 0.4, "n": 40, "lo": 0.2188, "hi": 0.6134, "ab": 0.0}, {"d": "2025-10-27", "acc": 0.4762, "n": 42, "lo": 0.2834, "hi": 0.6763, "ab": 0.0}, {"d": "2025-11-03", "acc": 0.6538, "n": 26, "lo": 0.3888, "hi": 0.8487, "ab": 0.0}, {"d": "2025-11-10", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.0}, {"d": "2025-11-17", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.4, "n": 30, "lo": 0.1982, "hi": 0.6425, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.4, "n": 30, "lo": 0.1982, "hi": 0.6425, "ab": 0.0}, {"d": "2025-12-08", "acc": 0.5, "n": 22, "lo": 0.2456, "hi": 0.7544, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.2895, "n": 38, "lo": 0.1356, "hi": 0.5142, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.4545, "n": 22, "lo": 0.2127, "hi": 0.7199, "ab": 0.0}, {"d": "2025-12-29", "acc": 0.4167, "n": 24, "lo": 0.1933, "hi": 0.6805, "ab": 0.0}, {"d": "2026-01-05", "acc": 0.1538, "n": 26, "lo": 0.0433, "hi": 0.4224, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.2812, "n": 32, "lo": 0.1213, "hi": 0.5259, "ab": 0.0}, {"d": "2026-01-19", "acc": 0.3529, "n": 34, "lo": 0.1731, "hi": 0.587, "ab": 0.0}, {"d": "2026-01-26", "acc": 0.3462, "n": 26, "lo": 0.1513, "hi": 0.6112, "ab": 0.0}, {"d": "2026-02-02", "acc": 0.25, "n": 32, "lo": 0.1018, "hi": 0.495, "ab": 0.0}, {"d": "2026-02-09", "acc": 0.4333, "n": 30, "lo": 0.2228, "hi": 0.6711, "ab": 0.0}, {"d": "2026-02-16", "acc": 0.1111, "n": 18, "lo": 0.0199, "hi": 0.435, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.3214, "n": 28, "lo": 0.1398, "hi": 0.58, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.4118, "n": 34, "lo": 0.2161, "hi": 0.6399, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.3571, "n": 42, "lo": 0.1895, "hi": 0.569, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.4, "n": 20, "lo": 0.1682, "hi": 0.6873, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.5312, "n": 32, "lo": 0.3055, "hi": 0.7449, "ab": 0.0312}, {"d": "2026-04-06", "acc": 0.5, "n": 34, "lo": 0.2853, "hi": 0.7147, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.3824, "n": 34, "lo": 0.1943, "hi": 0.6138, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.3, "n": 30, "lo": 0.1299, "hi": 0.5517, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.2105, "n": 38, "lo": 0.0851, "hi": 0.4333, "ab": 0.0}, {"d": "2026-05-04", "acc": 0.4706, "n": 34, "lo": 0.2616, "hi": 0.6904, "ab": 0.0}, {"d": "2026-05-11", "acc": 0.3529, "n": 34, "lo": 0.1731, "hi": 0.587, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.3077, "n": 26, "lo": 0.1268, "hi": 0.5763, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.0}, {"d": "2026-06-01", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2026-06-08", "acc": 0.4091, "n": 22, "lo": 0.1814, "hi": 0.6839, "ab": 0.0}, {"d": "2026-06-15", "acc": 0.3077, "n": 26, "lo": 0.1268, "hi": 0.5763, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.2308, "n": 26, "lo": 0.0818, "hi": 0.5026, "ab": 0.0}, {"d": "2026-06-29", "acc": 0.4722, "n": 36, "lo": 0.2677, "hi": 0.6865, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.1176, "n": 34, "lo": 0.0329, "hi": 0.3434, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.2105, "n": 38, "lo": 0.0851, "hi": 0.4333, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.3214, "n": 28, "lo": 0.1398, "hi": 0.58, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.3125, "n": 16, "lo": 0.1024, "hi": 0.6442, "ab": 0.0}], "samples": [{"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 99", "ok": true, "abstain": false, "conf": 99}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 99", "ok": true, "abstain": false, "conf": 99}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, a South American government became the third country in the world to recognise Edmundo González as the rightful winner of Venezuela's disputed presidential election. Which country was it?", "correct_opt": "Argentina", "model_opt": "Ecuador", "raw": "F, 84", "ok": false, "abstain": false, "conf": 84}, {"band": "mid", "date": "2025-06-17", "stem": "In mid-June 2025, the Florida Panthers wrapped up a six-game Stanley Cup Final to defend their title. Which team did they beat in the rematch of the previous year's final?", "correct_opt": "Edmonton Oilers", "model_opt": "Edmonton Oilers", "raw": "C, 99", "ok": true, "abstain": false, "conf": 99}, {"band": "mid", "date": "2025-06-20", "stem": "In late June 2025, Singapore signed a memorandum with a Southeast Asian neighbour providing for an annual exchange of between 300 and 1,000 skilled workers. Which country signed with Singapore?", "correct_opt": "Vietnam", "model_opt": "Cambodia", "raw": "B, 62", "ok": false, "abstain": false, "conf": 62}, {"band": "mid", "date": "2025-06-18", "stem": "In mid-June 2025, the U.S. Supreme Court split 6–3 to uphold a state statute barring gender-affirming medical treatment for minors, clearing the way for similar bans elsewhere. Which state's law was at issue?", "correct_opt": "Tennessee", "model_opt": "Tennessee", "raw": "F, 99", "ok": true, "abstain": false, "conf": 99}, {"band": "mid", "date": "2025-06-20", "stem": "In late June 2025, a European national parliament passed a resolution declaring the Soviet Union's 1944 mass deportation of the Crimean Tatars an act of genocide. Which country's parliament did so?", "correct_opt": "the Netherlands", "model_opt": "Belgium", "raw": "G, 63", "ok": false, "abstain": false, "conf": 63}, {"band": "post", "date": "2025-11-25", "stem": "In late November 2025, a long-dormant volcano in the Afar Region of Ethiopia erupted, spreading ash over the Arabian Peninsula and disrupting air traffic in northern India. Which volcano was it?", "correct_opt": "Hayli Gubbi", "model_opt": "Alu-Dalafilla", "raw": "G, 43", "ok": false, "abstain": false, "conf": 43}, {"band": "post", "date": "2025-11-25", "stem": "In late November 2025, Moldova reported that six Russian drones crossed into its airspace, one of them landing on the roof of a house. Near which town did that drone come down?", "correct_opt": "Florești", "model_opt": "Comrat", "raw": "A, 38", "ok": false, "abstain": false, "conf": 38}, {"band": "post", "date": "2025-11-25", "stem": "In late November 2025, Singapore ordered social media platforms to cut off local access to the accounts of a foreign preacher who urged replacing the city-state's secular order with an Islamic state. Of which country wa…", "correct_opt": "Australia", "model_opt": "Indonesia", "raw": "F, 58", "ok": false, "abstain": false, "conf": 58}]}, {"id": "gpt-5.4-mini", "name": "GPT-5.4 mini", "provider": "openai", "family": "gpt-5.4", "release": "2026-03-17", "pub_rel": "2025-08", "pub_train": "2025-08", "pub_rel_d": "2025-08-31", "pub_train_d": "2025-08-31", "fit": {"midpoint": "2025-12-10", "width": 2.97, "decay_start": "2025-12-03", "floor_reach": "2025-12-21", "floor": 0.3235, "plateau": 0.8297, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2025-12-06", "2025-12-15"]}, "weekly": [{"d": "2024-07-29", "acc": 0.8125, "n": 16, "lo": 0.4668, "hi": 0.9555, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.7308, "n": 26, "lo": 0.4598, "hi": 0.8964, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.625, "n": 24, "lo": 0.3544, "hi": 0.835, "ab": 0.0}, {"d": "2024-08-19", "acc": 0.7692, "n": 26, "lo": 0.4974, "hi": 0.9182, "ab": 0.0}, {"d": "2024-08-26", "acc": 0.5455, "n": 22, "lo": 0.2801, "hi": 0.7873, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.875, "n": 32, "lo": 0.6398, "hi": 0.965, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.6154, "n": 26, "lo": 0.3552, "hi": 0.8229, "ab": 0.0}, {"d": "2024-09-16", "acc": 0.7143, "n": 28, "lo": 0.4535, "hi": 0.8828, "ab": 0.0}, {"d": "2024-09-23", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2024-09-30", "acc": 0.875, "n": 24, "lo": 0.5975, "hi": 0.9706, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.8571, "n": 14, "lo": 0.4869, "hi": 0.9743, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.7727, "n": 22, "lo": 0.4775, "hi": 0.9267, "ab": 0.0}, {"d": "2024-10-21", "acc": 0.75, "n": 16, "lo": 0.4093, "hi": 0.9285, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.6667, "n": 24, "lo": 0.3906, "hi": 0.8619, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.7273, "n": 22, "lo": 0.4343, "hi": 0.9025, "ab": 0.0}, {"d": "2024-11-11", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.8333, "n": 18, "lo": 0.5067, "hi": 0.9605, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.7353, "n": 34, "lo": 0.4976, "hi": 0.8862, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.5, "n": 22, "lo": 0.2456, "hi": 0.7544, "ab": 0.0}, {"d": "2024-12-09", "acc": 0.9333, "n": 30, "lo": 0.7018, "hi": 0.9881, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.8421, "n": 38, "lo": 0.6243, "hi": 0.9448, "ab": 0.0}, {"d": "2024-12-23", "acc": 0.875, "n": 40, "lo": 0.6687, "hi": 0.9604, "ab": 0.0}, {"d": "2024-12-30", "acc": 1.0, "n": 10, "lo": 0.5655, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-06", "acc": 1.0, "n": 18, "lo": 0.7008, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-13", "acc": 0.75, "n": 16, "lo": 0.4093, "hi": 0.9285, "ab": 0.0}, {"d": "2025-01-20", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-27", "acc": 0.9, "n": 30, "lo": 0.6603, "hi": 0.9766, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.8571, "n": 28, "lo": 0.6006, "hi": 0.9599, "ab": 0.0}, {"d": "2025-02-10", "acc": 0.9545, "n": 22, "lo": 0.6785, "hi": 0.9952, "ab": 0.0}, {"d": "2025-02-17", "acc": 0.9, "n": 30, "lo": 0.6603, "hi": 0.9766, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.9643, "n": 28, "lo": 0.7324, "hi": 0.9963, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.75, "n": 32, "lo": 0.505, "hi": 0.8982, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.9444, "n": 36, "lo": 0.7424, "hi": 0.9901, "ab": 0.0}, {"d": "2025-03-17", "acc": 0.9231, "n": 26, "lo": 0.6669, "hi": 0.9863, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.95, "n": 20, "lo": 0.6555, "hi": 0.9948, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-04-07", "acc": 0.8056, "n": 36, "lo": 0.5773, "hi": 0.9263, "ab": 0.0}, {"d": "2025-04-14", "acc": 0.9545, "n": 22, "lo": 0.6785, "hi": 0.9952, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.6765, "n": 34, "lo": 0.4405, "hi": 0.8474, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.8929, "n": 28, "lo": 0.6417, "hi": 0.9749, "ab": 0.0}, {"d": "2025-05-05", "acc": 0.8214, "n": 28, "lo": 0.5615, "hi": 0.9429, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.8421, "n": 38, "lo": 0.6243, "hi": 0.9448, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.9667, "n": 30, "lo": 0.7465, "hi": 0.9965, "ab": 0.0}, {"d": "2025-05-26", "acc": 0.875, "n": 24, "lo": 0.5975, "hi": 0.9706, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.6471, "n": 34, "lo": 0.413, "hi": 0.8269, "ab": 0.0}, {"d": "2025-06-09", "acc": 0.8333, "n": 24, "lo": 0.552, "hi": 0.953, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.6333, "n": 30, "lo": 0.3869, "hi": 0.8254, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.8529, "n": 34, "lo": 0.6225, "hi": 0.9533, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.7105, "n": 38, "lo": 0.4858, "hi": 0.8644, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.875, "n": 24, "lo": 0.5975, "hi": 0.9706, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.6364, "n": 22, "lo": 0.3538, "hi": 0.8483, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.8333, "n": 30, "lo": 0.5839, "hi": 0.9469, "ab": 0.0}, {"d": "2025-07-28", "acc": 0.9, "n": 30, "lo": 0.6603, "hi": 0.9766, "ab": 0.0}, {"d": "2025-08-04", "acc": 0.9444, "n": 18, "lo": 0.6288, "hi": 0.9942, "ab": 0.0}, {"d": "2025-08-11", "acc": 0.8438, "n": 32, "lo": 0.6041, "hi": 0.9503, "ab": 0.0}, {"d": "2025-08-18", "acc": 0.9474, "n": 38, "lo": 0.7536, "hi": 0.9906, "ab": 0.0}, {"d": "2025-08-25", "acc": 0.881, "n": 42, "lo": 0.6817, "hi": 0.9624, "ab": 0.0}, {"d": "2025-09-01", "acc": 0.95, "n": 20, "lo": 0.6555, "hi": 0.9948, "ab": 0.0}, {"d": "2025-09-08", "acc": 1.0, "n": 28, "lo": 0.7847, "hi": 1.0, "ab": 0.0}, {"d": "2025-09-15", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2025-09-22", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2025-09-29", "acc": 0.7727, "n": 22, "lo": 0.4775, "hi": 0.9267, "ab": 0.0}, {"d": "2025-10-06", "acc": 0.8529, "n": 34, "lo": 0.6225, "hi": 0.9533, "ab": 0.0}, {"d": "2025-10-13", "acc": 0.8056, "n": 36, "lo": 0.5773, "hi": 0.9263, "ab": 0.0}, {"d": "2025-10-20", "acc": 0.85, "n": 40, "lo": 0.6396, "hi": 0.9476, "ab": 0.0}, {"d": "2025-10-27", "acc": 0.7381, "n": 42, "lo": 0.5245, "hi": 0.8781, "ab": 0.0}, {"d": "2025-11-03", "acc": 1.0, "n": 26, "lo": 0.7719, "hi": 1.0, "ab": 0.0}, {"d": "2025-11-10", "acc": 0.8929, "n": 28, "lo": 0.6417, "hi": 0.9749, "ab": 0.0}, {"d": "2025-11-17", "acc": 0.7727, "n": 22, "lo": 0.4775, "hi": 0.9267, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.7667, "n": 30, "lo": 0.5137, "hi": 0.9109, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.7667, "n": 30, "lo": 0.5137, "hi": 0.9109, "ab": 0.0}, {"d": "2025-12-08", "acc": 0.6364, "n": 22, "lo": 0.3538, "hi": 0.8483, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.3158, "n": 38, "lo": 0.1536, "hi": 0.5399, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.0}, {"d": "2025-12-29", "acc": 0.375, "n": 24, "lo": 0.165, "hi": 0.6456, "ab": 0.0}, {"d": "2026-01-05", "acc": 0.3462, "n": 26, "lo": 0.1513, "hi": 0.6112, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.375, "n": 32, "lo": 0.1848, "hi": 0.6136, "ab": 0.0}, {"d": "2026-01-19", "acc": 0.2941, "n": 34, "lo": 0.1328, "hi": 0.5313, "ab": 0.0}, {"d": "2026-01-26", "acc": 0.4231, "n": 26, "lo": 0.204, "hi": 0.6772, "ab": 0.0}, {"d": "2026-02-02", "acc": 0.3438, "n": 32, "lo": 0.1628, "hi": 0.5852, "ab": 0.0}, {"d": "2026-02-09", "acc": 0.4333, "n": 30, "lo": 0.2228, "hi": 0.6711, "ab": 0.0}, {"d": "2026-02-16", "acc": 0.0556, "n": 18, "lo": 0.0058, "hi": 0.3712, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.2857, "n": 28, "lo": 0.1172, "hi": 0.5465, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.4706, "n": 34, "lo": 0.2616, "hi": 0.6904, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.381, "n": 42, "lo": 0.2075, "hi": 0.5912, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.4, "n": 20, "lo": 0.1682, "hi": 0.6873, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.625, "n": 32, "lo": 0.3864, "hi": 0.8152, "ab": 0.0}, {"d": "2026-04-06", "acc": 0.3529, "n": 34, "lo": 0.1731, "hi": 0.587, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.2941, "n": 34, "lo": 0.1328, "hi": 0.5313, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.4333, "n": 30, "lo": 0.2228, "hi": 0.6711, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.2895, "n": 38, "lo": 0.1356, "hi": 0.5142, "ab": 0.0}, {"d": "2026-05-04", "acc": 0.4118, "n": 34, "lo": 0.2161, "hi": 0.6399, "ab": 0.0}, {"d": "2026-05-11", "acc": 0.3235, "n": 34, "lo": 0.1526, "hi": 0.5595, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.3462, "n": 26, "lo": 0.1513, "hi": 0.6112, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.0}, {"d": "2026-06-01", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.0}, {"d": "2026-06-08", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.0}, {"d": "2026-06-15", "acc": 0.2308, "n": 26, "lo": 0.0818, "hi": 0.5026, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.2692, "n": 26, "lo": 0.1036, "hi": 0.5402, "ab": 0.0}, {"d": "2026-06-29", "acc": 0.3611, "n": 36, "lo": 0.1826, "hi": 0.5885, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.1471, "n": 34, "lo": 0.0467, "hi": 0.3775, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.2368, "n": 38, "lo": 0.1012, "hi": 0.461, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.125, "n": 16, "lo": 0.0224, "hi": 0.4709, "ab": 0.0}], "samples": [{"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 98", "ok": true, "abstain": false, "conf": 98}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 98", "ok": true, "abstain": false, "conf": 98}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, a South American government became the third country in the world to recognise Edmundo González as the rightful winner of Venezuela's disputed presidential election. Which country was it?", "correct_opt": "Argentina", "model_opt": "Paraguay", "raw": "B, 72", "ok": false, "abstain": false, "conf": 72}, {"band": "mid", "date": "2025-12-04", "stem": "In early December 2025, the leader of an Israeli-backed anti-Hamas militia operating in southern Gaza was shot dead by unknown assailants in Rafah. Who was killed?", "correct_opt": "Yasser Abu Shabab", "model_opt": "Yasser Abu Shabab", "raw": "G, 82", "ok": true, "abstain": false, "conf": 82}, {"band": "mid", "date": "2025-12-04", "stem": "In early December 2025, the United States signed a five-year deal worth more than $1.6 billion to support one African country's health system, with the partner government pledging an extra $850 million of its own money.…", "correct_opt": "Kenya", "model_opt": "Zambia", "raw": "A, 55", "ok": false, "abstain": false, "conf": 55}, {"band": "mid", "date": "2025-12-04", "stem": "In early December 2025, four national broadcasters said they would sit out the next Eurovision Song Contest after the European Broadcasting Union confirmed Israel could compete. Alongside Ireland, Slovenia and the Nethe…", "correct_opt": "Spain", "model_opt": "Spain", "raw": "G, 78", "ok": true, "abstain": false, "conf": 78}, {"band": "mid", "date": "2025-12-04", "stem": "In early December 2025, Spanish police broke up a trafficking ring that had brought more than 300 labourers into the country on tourist visas and put them to work on farms. Most of the exploited workers came from which …", "correct_opt": "Nepal", "model_opt": "India", "raw": "D, 62", "ok": false, "abstain": false, "conf": 62}, {"band": "post", "date": "2025-12-22", "stem": "In late December 2025, China's commerce ministry announced provisional anti-dumping duties of 42.7% on one category of European Union agricultural exports. Which category was targeted?", "correct_opt": "Dairy products", "model_opt": "Brandy and wine spirits", "raw": "B, 62", "ok": false, "abstain": false, "conf": 62}, {"band": "post", "date": "2025-12-22", "stem": "In late December 2025, India concluded a free-trade agreement with a Pacific country that also promised US$20 billion of investment in the Indian economy over fifteen years. Which country was India's partner in this dea…", "correct_opt": "New Zealand", "model_opt": "Australia", "raw": "F, 84", "ok": false, "abstain": false, "conf": 84}, {"band": "post", "date": "2025-12-22", "stem": "In late December 2025, a driver plowed into spectators gathered for a parade in the province of Gelderland, injuring ten people. In which Dutch municipality did this happen?", "correct_opt": "Nunspeet", "model_opt": "Ede", "raw": "C, 22", "ok": false, "abstain": false, "conf": 22}]}, {"id": "gpt-5.4-nano", "name": "GPT-5.4 nano", "provider": "openai", "family": "gpt-5.4", "release": "2026-03-17", "pub_rel": "2025-08", "pub_train": "2025-08", "pub_rel_d": "2025-08-31", "pub_train_d": "2025-08-31", "fit": {"midpoint": "2025-02-23", "width": 5.72, "decay_start": "2025-02-11", "floor_reach": "2025-03-12", "floor": 0.1517, "plateau": 0.3037, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2025-02-11", "2025-03-12"]}, "weekly": [{"d": "2024-07-29", "acc": 0.5625, "n": 16, "lo": 0.2589, "hi": 0.8255, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.3846, "n": 26, "lo": 0.1771, "hi": 0.6448, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.25, "n": 24, "lo": 0.0889, "hi": 0.5323, "ab": 0.0}, {"d": "2024-08-19", "acc": 0.3077, "n": 26, "lo": 0.1268, "hi": 0.5763, "ab": 0.0}, {"d": "2024-08-26", "acc": 0.0909, "n": 22, "lo": 0.0162, "hi": 0.3774, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.3438, "n": 32, "lo": 0.1628, "hi": 0.5852, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.2308, "n": 26, "lo": 0.0818, "hi": 0.5026, "ab": 0.0}, {"d": "2024-09-16", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.0}, {"d": "2024-09-23", "acc": 0.5909, "n": 22, "lo": 0.3161, "hi": 0.8186, "ab": 0.0}, {"d": "2024-09-30", "acc": 0.4583, "n": 24, "lo": 0.2229, "hi": 0.714, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.1429, "n": 14, "lo": 0.0257, "hi": 0.5131, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.0}, {"d": "2024-10-21", "acc": 0.1875, "n": 16, "lo": 0.0445, "hi": 0.5332, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.2917, "n": 24, "lo": 0.1127, "hi": 0.5716, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.1818, "n": 22, "lo": 0.0514, "hi": 0.477, "ab": 0.0}, {"d": "2024-11-11", "acc": 0.25, "n": 24, "lo": 0.0889, "hi": 0.5323, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.3333, "n": 18, "lo": 0.1206, "hi": 0.6458, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.2941, "n": 34, "lo": 0.1328, "hi": 0.5313, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.0}, {"d": "2024-12-09", "acc": 0.3667, "n": 30, "lo": 0.1746, "hi": 0.6131, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.3158, "n": 38, "lo": 0.1536, "hi": 0.5399, "ab": 0.0}, {"d": "2024-12-23", "acc": 0.15, "n": 40, "lo": 0.0524, "hi": 0.3604, "ab": 0.0}, {"d": "2024-12-30", "acc": 0.2, "n": 10, "lo": 0.0362, "hi": 0.6245, "ab": 0.0}, {"d": "2025-01-06", "acc": 0.3333, "n": 18, "lo": 0.1206, "hi": 0.6458, "ab": 0.0}, {"d": "2025-01-13", "acc": 0.1875, "n": 16, "lo": 0.0445, "hi": 0.5332, "ab": 0.0}, {"d": "2025-01-20", "acc": 0.5, "n": 22, "lo": 0.2456, "hi": 0.7544, "ab": 0.0}, {"d": "2025-01-27", "acc": 0.3667, "n": 30, "lo": 0.1746, "hi": 0.6131, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.0}, {"d": "2025-02-10", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.0}, {"d": "2025-02-17", "acc": 0.2, "n": 30, "lo": 0.0705, "hi": 0.4519, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.1875, "n": 32, "lo": 0.0659, "hi": 0.4301, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.1667, "n": 36, "lo": 0.0584, "hi": 0.3922, "ab": 0.0}, {"d": "2025-03-17", "acc": 0.1154, "n": 26, "lo": 0.0271, "hi": 0.3791, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.2, "n": 20, "lo": 0.0567, "hi": 0.5098, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.1071, "n": 28, "lo": 0.0251, "hi": 0.3583, "ab": 0.0}, {"d": "2025-04-07", "acc": 0.0833, "n": 36, "lo": 0.0195, "hi": 0.2938, "ab": 0.0}, {"d": "2025-04-14", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.1471, "n": 34, "lo": 0.0467, "hi": 0.3775, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.0}, {"d": "2025-05-05", "acc": 0.1071, "n": 28, "lo": 0.0251, "hi": 0.3583, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.2368, "n": 38, "lo": 0.1012, "hi": 0.461, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.2, "n": 30, "lo": 0.0705, "hi": 0.4519, "ab": 0.0}, {"d": "2025-05-26", "acc": 0.1667, "n": 24, "lo": 0.047, "hi": 0.448, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.0588, "n": 34, "lo": 0.0105, "hi": 0.2698, "ab": 0.0}, {"d": "2025-06-09", "acc": 0.2083, "n": 24, "lo": 0.0669, "hi": 0.4912, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.2, "n": 30, "lo": 0.0705, "hi": 0.4519, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.0588, "n": 34, "lo": 0.0105, "hi": 0.2698, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.1579, "n": 38, "lo": 0.0552, "hi": 0.3757, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.0833, "n": 24, "lo": 0.0149, "hi": 0.3539, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.0455, "n": 22, "lo": 0.0048, "hi": 0.3215, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.2, "n": 30, "lo": 0.0705, "hi": 0.4519, "ab": 0.0}, {"d": "2025-07-28", "acc": 0.2, "n": 30, "lo": 0.0705, "hi": 0.4519, "ab": 0.0}, {"d": "2025-08-04", "acc": 0.3333, "n": 18, "lo": 0.1206, "hi": 0.6458, "ab": 0.0}, {"d": "2025-08-11", "acc": 0.1875, "n": 32, "lo": 0.0659, "hi": 0.4301, "ab": 0.0}, {"d": "2025-08-18", "acc": 0.1842, "n": 38, "lo": 0.0697, "hi": 0.4049, "ab": 0.0}, {"d": "2025-08-25", "acc": 0.1667, "n": 42, "lo": 0.0629, "hi": 0.3736, "ab": 0.0}, {"d": "2025-09-01", "acc": 0.1, "n": 20, "lo": 0.0179, "hi": 0.4042, "ab": 0.0}, {"d": "2025-09-08", "acc": 0.1429, "n": 28, "lo": 0.0401, "hi": 0.3994, "ab": 0.0}, {"d": "2025-09-15", "acc": 0.1818, "n": 22, "lo": 0.0514, "hi": 0.477, "ab": 0.0}, {"d": "2025-09-22", "acc": 0.1364, "n": 22, "lo": 0.0321, "hi": 0.4288, "ab": 0.0}, {"d": "2025-09-29", "acc": 0.0909, "n": 22, "lo": 0.0162, "hi": 0.3774, "ab": 0.0}, {"d": "2025-10-06", "acc": 0.2941, "n": 34, "lo": 0.1328, "hi": 0.5313, "ab": 0.0}, {"d": "2025-10-13", "acc": 0.1389, "n": 36, "lo": 0.0441, "hi": 0.3607, "ab": 0.0}, {"d": "2025-10-20", "acc": 0.075, "n": 40, "lo": 0.0175, "hi": 0.2694, "ab": 0.0}, {"d": "2025-10-27", "acc": 0.2143, "n": 42, "lo": 0.0912, "hi": 0.4258, "ab": 0.0}, {"d": "2025-11-03", "acc": 0.2308, "n": 26, "lo": 0.0818, "hi": 0.5026, "ab": 0.0}, {"d": "2025-11-10", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.0}, {"d": "2025-11-17", "acc": 0.0909, "n": 22, "lo": 0.0162, "hi": 0.3774, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.2, "n": 30, "lo": 0.0705, "hi": 0.4519, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.1, "n": 30, "lo": 0.0234, "hi": 0.3397, "ab": 0.0}, {"d": "2025-12-08", "acc": 0.2727, "n": 22, "lo": 0.0975, "hi": 0.5657, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.0789, "n": 38, "lo": 0.0184, "hi": 0.2811, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.0909, "n": 22, "lo": 0.0162, "hi": 0.3774, "ab": 0.0}, {"d": "2025-12-29", "acc": 0.125, "n": 24, "lo": 0.0294, "hi": 0.4025, "ab": 0.0}, {"d": "2026-01-05", "acc": 0.0769, "n": 26, "lo": 0.0137, "hi": 0.3331, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.1562, "n": 32, "lo": 0.0497, "hi": 0.3959, "ab": 0.0}, {"d": "2026-01-19", "acc": 0.0588, "n": 34, "lo": 0.0105, "hi": 0.2698, "ab": 0.0}, {"d": "2026-01-26", "acc": 0.0385, "n": 26, "lo": 0.004, "hi": 0.2835, "ab": 0.0}, {"d": "2026-02-02", "acc": 0.125, "n": 32, "lo": 0.035, "hi": 0.3602, "ab": 0.0}, {"d": "2026-02-09", "acc": 0.1333, "n": 30, "lo": 0.0374, "hi": 0.3788, "ab": 0.0}, {"d": "2026-02-16", "acc": 0.1111, "n": 18, "lo": 0.0199, "hi": 0.435, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.1429, "n": 28, "lo": 0.0401, "hi": 0.3994, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.3529, "n": 34, "lo": 0.1731, "hi": 0.587, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.1667, "n": 42, "lo": 0.0629, "hi": 0.3736, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.25, "n": 28, "lo": 0.0958, "hi": 0.5118, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.3, "n": 20, "lo": 0.1078, "hi": 0.6032, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.125, "n": 32, "lo": 0.035, "hi": 0.3602, "ab": 0.0}, {"d": "2026-04-06", "acc": 0.3235, "n": 34, "lo": 0.1526, "hi": 0.5595, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.0588, "n": 34, "lo": 0.0105, "hi": 0.2698, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.1, "n": 30, "lo": 0.0234, "hi": 0.3397, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.0526, "n": 38, "lo": 0.0094, "hi": 0.2464, "ab": 0.0}, {"d": "2026-05-04", "acc": 0.2941, "n": 34, "lo": 0.1328, "hi": 0.5313, "ab": 0.0}, {"d": "2026-05-11", "acc": 0.1176, "n": 34, "lo": 0.0329, "hi": 0.3434, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.1154, "n": 26, "lo": 0.0271, "hi": 0.3791, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.1071, "n": 28, "lo": 0.0251, "hi": 0.3583, "ab": 0.0}, {"d": "2026-06-01", "acc": 0.1429, "n": 28, "lo": 0.0401, "hi": 0.3994, "ab": 0.0}, {"d": "2026-06-08", "acc": 0.0909, "n": 22, "lo": 0.0162, "hi": 0.3774, "ab": 0.0}, {"d": "2026-06-15", "acc": 0.1923, "n": 26, "lo": 0.0616, "hi": 0.4634, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.0, "n": 26, "lo": 0.0, "hi": 0.2281, "ab": 0.0}, {"d": "2026-06-29", "acc": 0.1389, "n": 36, "lo": 0.0441, "hi": 0.3607, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.0588, "n": 34, "lo": 0.0105, "hi": 0.2698, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.1316, "n": 38, "lo": 0.0417, "hi": 0.3454, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.1786, "n": 28, "lo": 0.0571, "hi": 0.4385, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.0625, "n": 16, "lo": 0.0066, "hi": 0.4023, "ab": 0.0}], "samples": [{"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 88", "ok": true, "abstain": false, "conf": 88}, {"band": "pre", "date": "2024-08-03", "stem": "In early August 2024, Vietnam's ruling party formally installed a new General Secretary after the death of Nguyễn Phú Trọng the previous month. Who took the post?", "correct_opt": "Tô Lâm", "model_opt": "Tô Lâm", "raw": "C, 98", "ok": true, "abstain": false, "conf": 98}, {"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Qatar", "raw": "D, 60", "ok": false, "abstain": false, "conf": 60}, {"band": "mid", "date": "2025-02-15", "stem": "In mid-February 2025, African Union member states voted for a new chairperson of the AU Commission, replacing Chad's Moussa Faki. Who won the post?", "correct_opt": "Mahamoud Ali Youssouf of Djibouti", "model_opt": "Mahamoud Ali Youssouf of Djibouti", "raw": "F, 62", "ok": true, "abstain": false, "conf": 62}, {"band": "mid", "date": "2025-02-12", "stem": "In mid-February 2025, after the Romanian president's resignation took effect, a senior legislator assumed the duties of interim head of state. Who took over as Romania's acting president?", "correct_opt": "Ilie Bolojan", "model_opt": "Alina Gorghiu", "raw": "E, 55", "ok": false, "abstain": false, "conf": 55}, {"band": "mid", "date": "2025-02-13", "stem": "In mid-February 2025, a group of U.S. Justice Department lawyers, led by an acting federal prosecutor in Manhattan, resigned rather than carry out an instruction from Trump appointees. What were they ordered to do?", "correct_opt": "Dismiss the corruption prosecution of New York City Mayor Eric Adams", "model_opt": "Dismiss the corruption prosecution of New York City Mayor Eric Adams", "raw": "E, 72", "ok": true, "abstain": false, "conf": 72}, {"band": "mid", "date": "2025-02-13", "stem": "In mid-February 2025, a driver rammed a car into a crowd taking part in a demonstration in a southern German city, injuring dozens. Which organisation was holding the demonstration?", "correct_opt": "The ver.di trade union", "model_opt": "A Kurdish community association", "raw": "B, 55", "ok": false, "abstain": false, "conf": 55}, {"band": "post", "date": "2025-03-12", "stem": "In mid-March 2025, Switzerland's Federal Assembly filled the Federal Council vacancy left by the departure of the defence minister. Who was elected and given the defence portfolio?", "correct_opt": "Martin Pfister", "model_opt": "Elisabeth Schneider-Schneiter", "raw": "A, 62", "ok": false, "abstain": false, "conf": 62}, {"band": "post", "date": "2025-03-12", "stem": "In mid-March 2025, Canada announced counter-tariffs worth about CAD$29.8 billion on American goods in response to new U.S. duties on Canadian metals. Which Canadian minister made the announcement?", "correct_opt": "Dominic LeBlanc", "model_opt": "Mary Ng", "raw": "A, 55", "ok": false, "abstain": false, "conf": 55}, {"band": "post", "date": "2025-03-13", "stem": "In mid-March 2025, two Central Asian presidents signed an agreement definitively settling their countries' long-contested frontier, and transport links between them reopened. Which two countries reached the deal?", "correct_opt": "Kyrgyzstan and Tajikistan", "model_opt": "Kyrgyzstan and Uzbekistan", "raw": "E, 45", "ok": false, "abstain": false, "conf": 45}]}, {"id": "gpt-5.5", "name": "GPT-5.5", "provider": "openai", "family": "gpt-5.5", "release": "2026-04-23", "pub_rel": "2025-12", "pub_train": "2025-12", "pub_rel_d": "2025-12-31", "pub_train_d": "2025-12-31", "fit": {"midpoint": "2025-12-14", "width": 9.23, "decay_start": "2025-11-22", "floor_reach": "2026-01-14", "floor": 0.3758, "plateau": 0.9225, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2025-12-06", "2025-12-21"]}, "weekly": [{"d": "2024-07-29", "acc": 0.875, "n": 16, "lo": 0.5291, "hi": 0.9776, "ab": 0.0}, {"d": "2024-08-05", "acc": 1.0, "n": 26, "lo": 0.7719, "hi": 1.0, "ab": 0.0}, {"d": "2024-08-12", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2024-08-19", "acc": 0.9231, "n": 26, "lo": 0.6669, "hi": 0.9863, "ab": 0.0}, {"d": "2024-08-26", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2024-09-02", "acc": 1.0, "n": 32, "lo": 0.8064, "hi": 1.0, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.9231, "n": 26, "lo": 0.6669, "hi": 0.9863, "ab": 0.0}, {"d": "2024-09-16", "acc": 1.0, "n": 28, "lo": 0.7847, "hi": 1.0, "ab": 0.0}, {"d": "2024-09-23", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2024-09-30", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.8571, "n": 14, "lo": 0.4869, "hi": 0.9743, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2024-10-21", "acc": 1.0, "n": 16, "lo": 0.6756, "hi": 1.0, "ab": 0.0}, {"d": "2024-10-28", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2024-11-11", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2024-11-18", "acc": 1.0, "n": 18, "lo": 0.7008, "hi": 1.0, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.8529, "n": 34, "lo": 0.6225, "hi": 0.9533, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2024-12-09", "acc": 1.0, "n": 30, "lo": 0.7961, "hi": 1.0, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.9474, "n": 38, "lo": 0.7536, "hi": 0.9906, "ab": 0.0}, {"d": "2024-12-23", "acc": 1.0, "n": 40, "lo": 0.8389, "hi": 1.0, "ab": 0.0}, {"d": "2024-12-30", "acc": 1.0, "n": 10, "lo": 0.5655, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-06", "acc": 1.0, "n": 18, "lo": 0.7008, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-13", "acc": 1.0, "n": 16, "lo": 0.6756, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-20", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-27", "acc": 1.0, "n": 30, "lo": 0.7961, "hi": 1.0, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-02-10", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-02-17", "acc": 0.9667, "n": 30, "lo": 0.7465, "hi": 0.9965, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.8929, "n": 28, "lo": 0.6417, "hi": 0.9749, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.9375, "n": 32, "lo": 0.7167, "hi": 0.9889, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.8889, "n": 36, "lo": 0.672, "hi": 0.969, "ab": 0.0}, {"d": "2025-03-17", "acc": 0.9231, "n": 26, "lo": 0.6669, "hi": 0.9863, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.85, "n": 20, "lo": 0.5411, "hi": 0.9646, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-04-07", "acc": 0.9722, "n": 36, "lo": 0.7812, "hi": 0.9971, "ab": 0.0}, {"d": "2025-04-14", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.8235, "n": 34, "lo": 0.5897, "hi": 0.9381, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.8929, "n": 28, "lo": 0.6417, "hi": 0.9749, "ab": 0.0}, {"d": "2025-05-05", "acc": 1.0, "n": 28, "lo": 0.7847, "hi": 1.0, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.7895, "n": 38, "lo": 0.5667, "hi": 0.9149, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.9, "n": 30, "lo": 0.6603, "hi": 0.9766, "ab": 0.0}, {"d": "2025-05-26", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.7353, "n": 34, "lo": 0.4976, "hi": 0.8862, "ab": 0.0}, {"d": "2025-06-09", "acc": 0.9167, "n": 24, "lo": 0.6461, "hi": 0.9851, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.6333, "n": 30, "lo": 0.3869, "hi": 0.8254, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.9412, "n": 34, "lo": 0.7302, "hi": 0.9895, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.8421, "n": 38, "lo": 0.6243, "hi": 0.9448, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.9583, "n": 24, "lo": 0.6987, "hi": 0.9956, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.8333, "n": 30, "lo": 0.5839, "hi": 0.9469, "ab": 0.0}, {"d": "2025-07-28", "acc": 0.9, "n": 30, "lo": 0.6603, "hi": 0.9766, "ab": 0.0}, {"d": "2025-08-04", "acc": 1.0, "n": 18, "lo": 0.7008, "hi": 1.0, "ab": 0.0}, {"d": "2025-08-11", "acc": 0.9375, "n": 32, "lo": 0.7167, "hi": 0.9889, "ab": 0.0}, {"d": "2025-08-18", "acc": 0.9737, "n": 38, "lo": 0.7908, "hi": 0.9972, "ab": 0.0}, {"d": "2025-08-25", "acc": 0.9048, "n": 42, "lo": 0.7109, "hi": 0.9735, "ab": 0.0}, {"d": "2025-09-01", "acc": 1.0, "n": 20, "lo": 0.7225, "hi": 1.0, "ab": 0.0}, {"d": "2025-09-08", "acc": 1.0, "n": 28, "lo": 0.7847, "hi": 1.0, "ab": 0.0}, {"d": "2025-09-15", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-09-22", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2025-09-29", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2025-10-06", "acc": 0.8235, "n": 34, "lo": 0.5897, "hi": 0.9381, "ab": 0.0}, {"d": "2025-10-13", "acc": 0.9722, "n": 36, "lo": 0.7812, "hi": 0.9971, "ab": 0.0}, {"d": "2025-10-20", "acc": 0.925, "n": 40, "lo": 0.7306, "hi": 0.9825, "ab": 0.0}, {"d": "2025-10-27", "acc": 0.8095, "n": 42, "lo": 0.6, "hi": 0.9233, "ab": 0.0}, {"d": "2025-11-03", "acc": 1.0, "n": 26, "lo": 0.7719, "hi": 1.0, "ab": 0.0}, {"d": "2025-11-10", "acc": 1.0, "n": 28, "lo": 0.7847, "hi": 1.0, "ab": 0.0}, {"d": "2025-11-17", "acc": 0.7727, "n": 22, "lo": 0.4775, "hi": 0.9267, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.8333, "n": 30, "lo": 0.5839, "hi": 0.9469, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.8, "n": 30, "lo": 0.5481, "hi": 0.9295, "ab": 0.0}, {"d": "2025-12-08", "acc": 0.7727, "n": 22, "lo": 0.4775, "hi": 0.9267, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.6053, "n": 38, "lo": 0.3863, "hi": 0.7888, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.5, "n": 22, "lo": 0.2456, "hi": 0.7544, "ab": 0.0}, {"d": "2025-12-29", "acc": 0.4583, "n": 24, "lo": 0.2229, "hi": 0.714, "ab": 0.0}, {"d": "2026-01-05", "acc": 0.3846, "n": 26, "lo": 0.1771, "hi": 0.6448, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.25, "n": 32, "lo": 0.1018, "hi": 0.495, "ab": 0.0}, {"d": "2026-01-19", "acc": 0.3824, "n": 34, "lo": 0.1943, "hi": 0.6138, "ab": 0.0}, {"d": "2026-01-26", "acc": 0.4231, "n": 26, "lo": 0.204, "hi": 0.6772, "ab": 0.0}, {"d": "2026-02-02", "acc": 0.375, "n": 32, "lo": 0.1848, "hi": 0.6136, "ab": 0.0}, {"d": "2026-02-09", "acc": 0.6667, "n": 30, "lo": 0.4171, "hi": 0.8482, "ab": 0.0}, {"d": "2026-02-16", "acc": 0.2222, "n": 18, "lo": 0.0632, "hi": 0.5474, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.4643, "n": 28, "lo": 0.2404, "hi": 0.7035, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.4118, "n": 34, "lo": 0.2161, "hi": 0.6399, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.4048, "n": 42, "lo": 0.2259, "hi": 0.6131, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.2143, "n": 28, "lo": 0.0757, "hi": 0.4759, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.35, "n": 20, "lo": 0.1369, "hi": 0.6463, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.6562, "n": 32, "lo": 0.4148, "hi": 0.8372, "ab": 0.0}, {"d": "2026-04-06", "acc": 0.5882, "n": 34, "lo": 0.3601, "hi": 0.7839, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.4412, "n": 34, "lo": 0.2386, "hi": 0.6655, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.3667, "n": 30, "lo": 0.1746, "hi": 0.6131, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.2632, "n": 38, "lo": 0.1181, "hi": 0.4879, "ab": 0.0}, {"d": "2026-05-04", "acc": 0.4412, "n": 34, "lo": 0.2386, "hi": 0.6655, "ab": 0.0}, {"d": "2026-05-11", "acc": 0.3529, "n": 34, "lo": 0.1731, "hi": 0.587, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.4615, "n": 26, "lo": 0.2321, "hi": 0.7086, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.2857, "n": 28, "lo": 0.1172, "hi": 0.5465, "ab": 0.0}, {"d": "2026-06-01", "acc": 0.2857, "n": 28, "lo": 0.1172, "hi": 0.5465, "ab": 0.0}, {"d": "2026-06-08", "acc": 0.3636, "n": 22, "lo": 0.1517, "hi": 0.6462, "ab": 0.0}, {"d": "2026-06-15", "acc": 0.5, "n": 26, "lo": 0.2612, "hi": 0.7388, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.1923, "n": 26, "lo": 0.0616, "hi": 0.4634, "ab": 0.0}, {"d": "2026-06-29", "acc": 0.4167, "n": 36, "lo": 0.224, "hi": 0.6386, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.2059, "n": 34, "lo": 0.0782, "hi": 0.442, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.2632, "n": 38, "lo": 0.1181, "hi": 0.4879, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.125, "n": 16, "lo": 0.0224, "hi": 0.4709, "ab": 0.0}], "samples": [{"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 95", "ok": true, "abstain": false, "conf": 95}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 98", "ok": true, "abstain": false, "conf": 98}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, a South American government became the third country in the world to recognise Edmundo González as the rightful winner of Venezuela's disputed presidential election. Which country was it?", "correct_opt": "Argentina", "model_opt": "Uruguay", "raw": "C, 70", "ok": false, "abstain": false, "conf": 70}, {"band": "mid", "date": "2025-11-24", "stem": "In late November 2025, gunmen ambushed a governor's convoy on a road linking a southwestern Yemeni city to the rest of the country, killing about seven people including two of the attackers. Which governorate's governor…", "correct_opt": "Taiz", "model_opt": "Taiz", "raw": "G, 85", "ok": true, "abstain": false, "conf": 85}, {"band": "mid", "date": "2025-11-24", "stem": "In late November 2025, a suicide bomber struck outside the Federal Constabulary headquarters in Peshawar, Pakistan, killing three people before two more gunmen tried to storm the compound. Which militant group claimed t…", "correct_opt": "Jamaat-ul-Ahrar", "model_opt": "Hizbul Ahrar", "raw": "G, 72", "ok": false, "abstain": false, "conf": 72}, {"band": "mid", "date": "2025-11-24", "stem": "In late November 2025, a U.S. federal judge threw out indictments against James Comey and Letitia James after finding that a prosecutor had been unlawfully installed. Which prosecutor's appointment was ruled invalid?", "correct_opt": "Lindsey Halligan", "model_opt": "Lindsey Halligan", "raw": "D, 95", "ok": true, "abstain": false, "conf": 95}, {"band": "mid", "date": "2025-11-24", "stem": "In late November 2025, seven militants, among them a field commander, were killed in a clash tied to a clan feud in Cotabato in the southern Philippines. Which armed organisation did the slain commander belong to?", "correct_opt": "Moro Islamic Liberation Front", "model_opt": "Bangsamoro Islamic Freedom Fighters", "raw": "E, 75", "ok": false, "abstain": false, "conf": 75}, {"band": "post", "date": "2026-01-14", "stem": "In mid-January 2026, two Swiss cantons prohibited pyrotechnics inside enclosed venues after a nightclub blaze in an Alpine resort killed 40 people. In which resort did the deadly fire occur?", "correct_opt": "Crans-Montana", "model_opt": "Verbier", "raw": "H, 25", "ok": false, "abstain": false, "conf": 25}, {"band": "post", "date": "2026-01-15", "stem": "In mid-January 2026, the U.S. Coast Guard boarded and took control of a sanctioned tanker in the Caribbean during an American operation against oil shipments. Under which country's flag was the tanker sailing?", "correct_opt": "Guyana", "model_opt": "Panama", "raw": "F, 35", "ok": false, "abstain": false, "conf": 35}, {"band": "post", "date": "2026-01-15", "stem": "In mid-January 2026, a blast believed to have been caused by gas injured several people and brought down buildings in the centre of a Dutch city. Which city was hit?", "correct_opt": "Utrecht", "model_opt": "Rotterdam", "raw": "H, 35", "ok": false, "abstain": false, "conf": 35}]}, {"id": "gpt-5.6-luna", "name": "GPT-5.6 Luna", "provider": "openai", "family": "gpt-5.6", "release": "2026-07-09", "pub_rel": "2026-02", "pub_train": "2026-02", "pub_rel_d": "2026-02-28", "pub_train_d": "2026-02-28", "fit": {"midpoint": "2026-04-17", "width": 113.04, "decay_start": "2025-05-19", "floor_reach": null, "floor": 0.0, "plateau": 0.8692, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2026-03-09", "2026-04-29"]}, "weekly": [{"d": "2024-07-29", "acc": 0.875, "n": 16, "lo": 0.5291, "hi": 0.9776, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.9615, "n": 26, "lo": 0.7165, "hi": 0.996, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.9167, "n": 24, "lo": 0.6461, "hi": 0.9851, "ab": 0.0}, {"d": "2024-08-19", "acc": 0.7692, "n": 26, "lo": 0.4974, "hi": 0.9182, "ab": 0.0}, {"d": "2024-08-26", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.9688, "n": 32, "lo": 0.7593, "hi": 0.9967, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.6538, "n": 26, "lo": 0.3888, "hi": 0.8487, "ab": 0.0}, {"d": "2024-09-16", "acc": 0.8929, "n": 28, "lo": 0.6417, "hi": 0.9749, "ab": 0.0}, {"d": "2024-09-23", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2024-09-30", "acc": 0.9167, "n": 24, "lo": 0.6461, "hi": 0.9851, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.9286, "n": 14, "lo": 0.5609, "hi": 0.9925, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2024-10-21", "acc": 0.6875, "n": 16, "lo": 0.3558, "hi": 0.8976, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.8333, "n": 24, "lo": 0.552, "hi": 0.953, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2024-11-11", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.9444, "n": 18, "lo": 0.6288, "hi": 0.9942, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.8235, "n": 34, "lo": 0.5897, "hi": 0.9381, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.6818, "n": 22, "lo": 0.3932, "hi": 0.8764, "ab": 0.0}, {"d": "2024-12-09", "acc": 0.9, "n": 30, "lo": 0.6603, "hi": 0.9766, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.8684, "n": 38, "lo": 0.6546, "hi": 0.9583, "ab": 0.0}, {"d": "2024-12-23", "acc": 0.875, "n": 40, "lo": 0.6687, "hi": 0.9604, "ab": 0.0}, {"d": "2024-12-30", "acc": 1.0, "n": 10, "lo": 0.5655, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-06", "acc": 1.0, "n": 18, "lo": 0.7008, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-13", "acc": 0.875, "n": 16, "lo": 0.5291, "hi": 0.9776, "ab": 0.0}, {"d": "2025-01-20", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-27", "acc": 0.9333, "n": 30, "lo": 0.7018, "hi": 0.9881, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-02-10", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-02-17", "acc": 0.7667, "n": 30, "lo": 0.5137, "hi": 0.9109, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.875, "n": 32, "lo": 0.6398, "hi": 0.965, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.8889, "n": 36, "lo": 0.672, "hi": 0.969, "ab": 0.0}, {"d": "2025-03-17", "acc": 0.9615, "n": 26, "lo": 0.7165, "hi": 0.996, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.75, "n": 20, "lo": 0.4422, "hi": 0.9191, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.9643, "n": 28, "lo": 0.7324, "hi": 0.9963, "ab": 0.0}, {"d": "2025-04-07", "acc": 0.75, "n": 36, "lo": 0.5192, "hi": 0.8929, "ab": 0.0}, {"d": "2025-04-14", "acc": 0.6818, "n": 22, "lo": 0.3932, "hi": 0.8764, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.6471, "n": 34, "lo": 0.413, "hi": 0.8269, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.8571, "n": 28, "lo": 0.6006, "hi": 0.9599, "ab": 0.0}, {"d": "2025-05-05", "acc": 0.7143, "n": 28, "lo": 0.4535, "hi": 0.8828, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.7105, "n": 38, "lo": 0.4858, "hi": 0.8644, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.9, "n": 30, "lo": 0.6603, "hi": 0.9766, "ab": 0.0}, {"d": "2025-05-26", "acc": 0.875, "n": 24, "lo": 0.5975, "hi": 0.9706, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.7647, "n": 34, "lo": 0.5274, "hi": 0.9045, "ab": 0.0}, {"d": "2025-06-09", "acc": 0.7083, "n": 24, "lo": 0.4284, "hi": 0.8873, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.6, "n": 30, "lo": 0.3575, "hi": 0.8018, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.8235, "n": 34, "lo": 0.5897, "hi": 0.9381, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.6053, "n": 38, "lo": 0.3863, "hi": 0.7888, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.7917, "n": 24, "lo": 0.5088, "hi": 0.9331, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.5455, "n": 22, "lo": 0.2801, "hi": 0.7873, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.5667, "n": 30, "lo": 0.3289, "hi": 0.7772, "ab": 0.0}, {"d": "2025-07-28", "acc": 0.7667, "n": 30, "lo": 0.5137, "hi": 0.9109, "ab": 0.0}, {"d": "2025-08-04", "acc": 0.8889, "n": 18, "lo": 0.565, "hi": 0.9801, "ab": 0.0}, {"d": "2025-08-11", "acc": 0.4375, "n": 32, "lo": 0.231, "hi": 0.6682, "ab": 0.0}, {"d": "2025-08-18", "acc": 0.8158, "n": 38, "lo": 0.5951, "hi": 0.9303, "ab": 0.0}, {"d": "2025-08-25", "acc": 0.8571, "n": 42, "lo": 0.6536, "hi": 0.9502, "ab": 0.0}, {"d": "2025-09-01", "acc": 1.0, "n": 20, "lo": 0.7225, "hi": 1.0, "ab": 0.0}, {"d": "2025-09-08", "acc": 1.0, "n": 28, "lo": 0.7847, "hi": 1.0, "ab": 0.0}, {"d": "2025-09-15", "acc": 0.7273, "n": 22, "lo": 0.4343, "hi": 0.9025, "ab": 0.0}, {"d": "2025-09-22", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2025-09-29", "acc": 0.6364, "n": 22, "lo": 0.3538, "hi": 0.8483, "ab": 0.0}, {"d": "2025-10-06", "acc": 0.8235, "n": 34, "lo": 0.5897, "hi": 0.9381, "ab": 0.0}, {"d": "2025-10-13", "acc": 0.9167, "n": 36, "lo": 0.7062, "hi": 0.9805, "ab": 0.0}, {"d": "2025-10-20", "acc": 0.6, "n": 40, "lo": 0.3866, "hi": 0.7812, "ab": 0.0}, {"d": "2025-10-27", "acc": 0.8095, "n": 42, "lo": 0.6, "hi": 0.9233, "ab": 0.0}, {"d": "2025-11-03", "acc": 0.8846, "n": 26, "lo": 0.6209, "hi": 0.9729, "ab": 0.0}, {"d": "2025-11-10", "acc": 0.7857, "n": 28, "lo": 0.5241, "hi": 0.9243, "ab": 0.0}, {"d": "2025-11-17", "acc": 0.7727, "n": 22, "lo": 0.4775, "hi": 0.9267, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.7333, "n": 30, "lo": 0.4805, "hi": 0.891, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.6667, "n": 30, "lo": 0.4171, "hi": 0.8482, "ab": 0.0}, {"d": "2025-12-08", "acc": 0.7273, "n": 22, "lo": 0.4343, "hi": 0.9025, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.6579, "n": 38, "lo": 0.435, "hi": 0.8277, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2025-12-29", "acc": 0.7917, "n": 24, "lo": 0.5088, "hi": 0.9331, "ab": 0.0}, {"d": "2026-01-05", "acc": 0.4615, "n": 26, "lo": 0.2321, "hi": 0.7086, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.5312, "n": 32, "lo": 0.3055, "hi": 0.7449, "ab": 0.0}, {"d": "2026-01-19", "acc": 0.5882, "n": 34, "lo": 0.3601, "hi": 0.7839, "ab": 0.0}, {"d": "2026-01-26", "acc": 0.5385, "n": 26, "lo": 0.2914, "hi": 0.7679, "ab": 0.0}, {"d": "2026-02-02", "acc": 0.4375, "n": 32, "lo": 0.231, "hi": 0.6682, "ab": 0.0}, {"d": "2026-02-09", "acc": 0.5667, "n": 30, "lo": 0.3289, "hi": 0.7772, "ab": 0.0}, {"d": "2026-02-16", "acc": 0.4444, "n": 18, "lo": 0.1888, "hi": 0.7334, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.3929, "n": 28, "lo": 0.1881, "hi": 0.6437, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.5588, "n": 34, "lo": 0.3345, "hi": 0.7614, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.5476, "n": 42, "lo": 0.3444, "hi": 0.7361, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.5, "n": 20, "lo": 0.2366, "hi": 0.7634, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.5, "n": 32, "lo": 0.28, "hi": 0.72, "ab": 0.0}, {"d": "2026-04-06", "acc": 0.5588, "n": 34, "lo": 0.3345, "hi": 0.7614, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.5, "n": 34, "lo": 0.2853, "hi": 0.7147, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.6667, "n": 30, "lo": 0.4171, "hi": 0.8482, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.2632, "n": 38, "lo": 0.1181, "hi": 0.4879, "ab": 0.0}, {"d": "2026-05-04", "acc": 0.3824, "n": 34, "lo": 0.1943, "hi": 0.6138, "ab": 0.0}, {"d": "2026-05-11", "acc": 0.4118, "n": 34, "lo": 0.2161, "hi": 0.6399, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.3846, "n": 26, "lo": 0.1771, "hi": 0.6448, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.0}, {"d": "2026-06-01", "acc": 0.3214, "n": 28, "lo": 0.1398, "hi": 0.58, "ab": 0.0}, {"d": "2026-06-08", "acc": 0.3636, "n": 22, "lo": 0.1517, "hi": 0.6462, "ab": 0.0}, {"d": "2026-06-15", "acc": 0.3462, "n": 26, "lo": 0.1513, "hi": 0.6112, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.3462, "n": 26, "lo": 0.1513, "hi": 0.6112, "ab": 0.0}, {"d": "2026-06-29", "acc": 0.2778, "n": 36, "lo": 0.125, "hi": 0.5087, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.1471, "n": 34, "lo": 0.0467, "hi": 0.3775, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.1579, "n": 38, "lo": 0.0552, "hi": 0.3757, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.3214, "n": 28, "lo": 0.1398, "hi": 0.58, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.25, "n": 16, "lo": 0.0715, "hi": 0.5907, "ab": 0.0}], "samples": [{"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 99", "ok": true, "abstain": false, "conf": 99}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 100", "ok": true, "abstain": false, "conf": 100}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, a South American government became the third country in the world to recognise Edmundo González as the rightful winner of Venezuela's disputed presidential election. Which country was it?", "correct_opt": "Argentina", "model_opt": "Costa Rica", "raw": "D, 95", "ok": false, "abstain": false, "conf": 95}, {"band": "mid", "date": "2025-05-20", "stem": "In mid-May 2025, an Indian writer and her translator shared the International Booker Prize for a collection of short stories. What was the title of the winning book?", "correct_opt": "Heart Lamp", "model_opt": "Heart Lamp", "raw": "E, 100", "ok": true, "abstain": false, "conf": 100}, {"band": "mid", "date": "2025-05-23", "stem": "In late May 2025, a Georgian court ordered pre-trial detention for a leader of the Coalition for Change after he refused to appear before a parliamentary inquiry. Which opposition figure was jailed?", "correct_opt": "Zurab Japaridze", "model_opt": "Nika Gvaramia", "raw": "G, 98", "ok": false, "abstain": false, "conf": 98}, {"band": "mid", "date": "2025-05-20", "stem": "In mid-May 2025, three pupils were wounded in a knife attack at a school by a 16-year-old who had published a manifesto declaring women his intended targets. In which Finnish municipality did the attack take place?", "correct_opt": "Pirkkala", "model_opt": "Pirkkala", "raw": "B, 98", "ok": true, "abstain": false, "conf": 98}, {"band": "mid", "date": "2025-05-28", "stem": "In late May 2025, two Malaysian cabinet ministers said they would quit the government after losing internal contests in the People's Justice Party. Which pair announced their resignations?", "correct_opt": "Rafizi Ramli and Nik Nazmi", "model_opt": "Rafizi Ramli and Zaliha Mustafa", "raw": "B, 98", "ok": false, "abstain": false, "conf": 98}]}, {"id": "gpt-5.6-sol", "name": "GPT-5.6 Sol", "provider": "openai", "family": "gpt-5.6", "release": "2026-07-09", "pub_rel": "2026-02", "pub_train": "2026-02", "pub_rel_d": "2026-02-28", "pub_train_d": "2026-02-28", "fit": {"midpoint": "2026-03-05", "width": 22.58, "decay_start": "2026-01-10", "floor_reach": "2026-05-19", "floor": 0.3923, "plateau": 0.9551, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2026-02-15", "2026-03-20"]}, "weekly": [{"d": "2024-07-29", "acc": 1.0, "n": 16, "lo": 0.6756, "hi": 1.0, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.9615, "n": 26, "lo": 0.7165, "hi": 0.996, "ab": 0.0}, {"d": "2024-08-12", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2024-08-19", "acc": 0.8846, "n": 26, "lo": 0.6209, "hi": 0.9729, "ab": 0.0}, {"d": "2024-08-26", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2024-09-02", "acc": 1.0, "n": 32, "lo": 0.8064, "hi": 1.0, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.9615, "n": 26, "lo": 0.7165, "hi": 0.996, "ab": 0.0}, {"d": "2024-09-16", "acc": 1.0, "n": 28, "lo": 0.7847, "hi": 1.0, "ab": 0.0}, {"d": "2024-09-23", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2024-09-30", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.8571, "n": 14, "lo": 0.4869, "hi": 0.9743, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2024-10-21", "acc": 1.0, "n": 16, "lo": 0.6756, "hi": 1.0, "ab": 0.0}, {"d": "2024-10-28", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2024-11-04", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2024-11-11", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.8889, "n": 18, "lo": 0.565, "hi": 0.9801, "ab": 0.0}, {"d": "2024-11-25", "acc": 1.0, "n": 34, "lo": 0.8157, "hi": 1.0, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2024-12-09", "acc": 1.0, "n": 30, "lo": 0.7961, "hi": 1.0, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.9474, "n": 38, "lo": 0.7536, "hi": 0.9906, "ab": 0.0}, {"d": "2024-12-23", "acc": 1.0, "n": 40, "lo": 0.8389, "hi": 1.0, "ab": 0.0}, {"d": "2024-12-30", "acc": 1.0, "n": 10, "lo": 0.5655, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-06", "acc": 1.0, "n": 18, "lo": 0.7008, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-13", "acc": 1.0, "n": 16, "lo": 0.6756, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-20", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-27", "acc": 1.0, "n": 30, "lo": 0.7961, "hi": 1.0, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-02-10", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-02-17", "acc": 1.0, "n": 30, "lo": 0.7961, "hi": 1.0, "ab": 0.0}, {"d": "2025-02-24", "acc": 1.0, "n": 28, "lo": 0.7847, "hi": 1.0, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.875, "n": 32, "lo": 0.6398, "hi": 0.965, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.9444, "n": 36, "lo": 0.7424, "hi": 0.9901, "ab": 0.0}, {"d": "2025-03-17", "acc": 1.0, "n": 26, "lo": 0.7719, "hi": 1.0, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.9, "n": 20, "lo": 0.5958, "hi": 0.9821, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-04-07", "acc": 1.0, "n": 36, "lo": 0.8241, "hi": 1.0, "ab": 0.0}, {"d": "2025-04-14", "acc": 0.8636, "n": 22, "lo": 0.5712, "hi": 0.9679, "ab": 0.0}, {"d": "2025-04-21", "acc": 1.0, "n": 34, "lo": 0.8157, "hi": 1.0, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-05-05", "acc": 1.0, "n": 28, "lo": 0.7847, "hi": 1.0, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.8158, "n": 38, "lo": 0.5951, "hi": 0.9303, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.9667, "n": 30, "lo": 0.7465, "hi": 0.9965, "ab": 0.0}, {"d": "2025-05-26", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.7941, "n": 34, "lo": 0.558, "hi": 0.9218, "ab": 0.0}, {"d": "2025-06-09", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.7667, "n": 30, "lo": 0.5137, "hi": 0.9109, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.9118, "n": 34, "lo": 0.6924, "hi": 0.9794, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.8421, "n": 38, "lo": 0.6243, "hi": 0.9448, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.875, "n": 24, "lo": 0.5975, "hi": 0.9706, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.8667, "n": 30, "lo": 0.6212, "hi": 0.9626, "ab": 0.0}, {"d": "2025-07-28", "acc": 1.0, "n": 30, "lo": 0.7961, "hi": 1.0, "ab": 0.0}, {"d": "2025-08-04", "acc": 1.0, "n": 18, "lo": 0.7008, "hi": 1.0, "ab": 0.0}, {"d": "2025-08-11", "acc": 1.0, "n": 32, "lo": 0.8064, "hi": 1.0, "ab": 0.0}, {"d": "2025-08-18", "acc": 1.0, "n": 38, "lo": 0.8318, "hi": 1.0, "ab": 0.0}, {"d": "2025-08-25", "acc": 0.9524, "n": 42, "lo": 0.7733, "hi": 0.9915, "ab": 0.0}, {"d": "2025-09-01", "acc": 1.0, "n": 20, "lo": 0.7225, "hi": 1.0, "ab": 0.0}, {"d": "2025-09-08", "acc": 1.0, "n": 28, "lo": 0.7847, "hi": 1.0, "ab": 0.0}, {"d": "2025-09-15", "acc": 0.9545, "n": 22, "lo": 0.6785, "hi": 0.9952, "ab": 0.0}, {"d": "2025-09-22", "acc": 0.9545, "n": 22, "lo": 0.6785, "hi": 0.9952, "ab": 0.0}, {"d": "2025-09-29", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-10-06", "acc": 1.0, "n": 34, "lo": 0.8157, "hi": 1.0, "ab": 0.0}, {"d": "2025-10-13", "acc": 0.9167, "n": 36, "lo": 0.7062, "hi": 0.9805, "ab": 0.0}, {"d": "2025-10-20", "acc": 0.95, "n": 40, "lo": 0.7639, "hi": 0.9911, "ab": 0.0}, {"d": "2025-10-27", "acc": 0.9762, "n": 42, "lo": 0.8076, "hi": 0.9975, "ab": 0.0}, {"d": "2025-11-03", "acc": 1.0, "n": 26, "lo": 0.7719, "hi": 1.0, "ab": 0.0}, {"d": "2025-11-10", "acc": 1.0, "n": 28, "lo": 0.7847, "hi": 1.0, "ab": 0.0}, {"d": "2025-11-17", "acc": 0.9545, "n": 22, "lo": 0.6785, "hi": 0.9952, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.9333, "n": 30, "lo": 0.7018, "hi": 0.9881, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.9333, "n": 30, "lo": 0.7018, "hi": 0.9881, "ab": 0.0}, {"d": "2025-12-08", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.8947, "n": 38, "lo": 0.6861, "hi": 0.9706, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.9545, "n": 22, "lo": 0.6785, "hi": 0.9952, "ab": 0.0}, {"d": "2025-12-29", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2026-01-05", "acc": 0.9615, "n": 26, "lo": 0.7165, "hi": 0.996, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.9062, "n": 32, "lo": 0.6771, "hi": 0.978, "ab": 0.0}, {"d": "2026-01-19", "acc": 0.8235, "n": 34, "lo": 0.5897, "hi": 0.9381, "ab": 0.0}, {"d": "2026-01-26", "acc": 0.8462, "n": 26, "lo": 0.5776, "hi": 0.9567, "ab": 0.0}, {"d": "2026-02-02", "acc": 0.8438, "n": 32, "lo": 0.6041, "hi": 0.9503, "ab": 0.0}, {"d": "2026-02-09", "acc": 0.9667, "n": 30, "lo": 0.7465, "hi": 0.9965, "ab": 0.0}, {"d": "2026-02-16", "acc": 0.7778, "n": 18, "lo": 0.4526, "hi": 0.9368, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.5714, "n": 28, "lo": 0.3259, "hi": 0.7862, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.5588, "n": 34, "lo": 0.3345, "hi": 0.7614, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.619, "n": 42, "lo": 0.4088, "hi": 0.7925, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.3929, "n": 28, "lo": 0.1881, "hi": 0.6437, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.65, "n": 20, "lo": 0.3537, "hi": 0.8631, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.6875, "n": 32, "lo": 0.444, "hi": 0.8584, "ab": 0.0}, {"d": "2026-04-06", "acc": 0.6765, "n": 34, "lo": 0.4405, "hi": 0.8474, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.5588, "n": 34, "lo": 0.3345, "hi": 0.7614, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.4667, "n": 30, "lo": 0.2481, "hi": 0.6988, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.3158, "n": 38, "lo": 0.1536, "hi": 0.5399, "ab": 0.0}, {"d": "2026-05-04", "acc": 0.4118, "n": 34, "lo": 0.2161, "hi": 0.6399, "ab": 0.0}, {"d": "2026-05-11", "acc": 0.5, "n": 34, "lo": 0.2853, "hi": 0.7147, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.4231, "n": 26, "lo": 0.204, "hi": 0.6772, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.4286, "n": 28, "lo": 0.2138, "hi": 0.6741, "ab": 0.0}, {"d": "2026-06-01", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.0}, {"d": "2026-06-08", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.0}, {"d": "2026-06-15", "acc": 0.3846, "n": 26, "lo": 0.1771, "hi": 0.6448, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.3077, "n": 26, "lo": 0.1268, "hi": 0.5763, "ab": 0.0}, {"d": "2026-06-29", "acc": 0.4167, "n": 36, "lo": 0.224, "hi": 0.6386, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.2353, "n": 34, "lo": 0.0955, "hi": 0.4726, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.4474, "n": 38, "lo": 0.2521, "hi": 0.6603, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.5357, "n": 28, "lo": 0.2965, "hi": 0.7596, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.3125, "n": 16, "lo": 0.1024, "hi": 0.6442, "ab": 0.0}], "samples": [{"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 99", "ok": true, "abstain": false, "conf": 99}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 99", "ok": true, "abstain": false, "conf": 99}, {"band": "pre", "date": "2024-08-21", "stem": "In late August 2024, several suspected infections with the newer mpox strain were reported in a rural county of southwest England. Which county was involved?", "correct_opt": "Somerset", "model_opt": "Gloucestershire", "raw": "D, 45", "ok": false, "abstain": false, "conf": 45}, {"band": "mid", "date": "2026-01-12", "stem": "In mid-January 2026, a former prime minister of Georgia admitted guilt on large-scale money-laundering charges after investigators seized millions of dollars in cash during searches of his home. Who was this former prim…", "correct_opt": "Irakli Garibashvili", "model_opt": "Irakli Garibashvili", "raw": "D, 95", "ok": true, "abstain": false, "conf": 95}, {"band": "mid", "date": "2026-01-12", "stem": "In mid-January 2026, Swedish police announced their country's largest ever single cocaine seizure, three tonnes hidden in a container arriving from South America. At which port was the container found?", "correct_opt": "Helsingborg", "model_opt": "Gothenburg", "raw": "H, 85", "ok": false, "abstain": false, "conf": 85}, {"band": "mid", "date": "2026-01-12", "stem": "In mid-January 2026, a former leader of a Corsican nationalist armed movement was shot dead by a gunman while attending his mother's funeral. What was the victim's name?", "correct_opt": "Alain Orsoni", "model_opt": "Alain Orsoni", "raw": "B, 98", "ok": true, "abstain": false, "conf": 98}, {"band": "mid", "date": "2026-01-19", "stem": "In mid-January 2026, a Pacific island nation and New Zealand signed a wide-ranging partnership covering economic opportunity, foreign policy, health and security — their first bilateral deal since relations soured the p…", "correct_opt": "Kiribati", "model_opt": "Cook Islands", "raw": "A, 98", "ok": false, "abstain": false, "conf": 98}, {"band": "post", "date": "2026-05-19", "stem": "In mid-May 2026, Estonia said a NATO fighter jet had destroyed a drone above its territory, an aircraft believed to have been Ukrainian and thrown off course by Russian jamming. Which country's air force flew the jet th…", "correct_opt": "Romania", "model_opt": "Portugal", "raw": "F, 35", "ok": false, "abstain": false, "conf": 35}, {"band": "post", "date": "2026-05-19", "stem": "In mid-May 2026, a war powers measure aimed at halting unauthorised hostilities against Iran cleared a preliminary vote in the U.S. Senate after a single Republican broke ranks. Which senator's defection allowed it to a…", "correct_opt": "Bill Cassidy of Louisiana", "model_opt": "Rand Paul of Kentucky", "raw": "B, 92", "ok": false, "abstain": false, "conf": 92}, {"band": "post", "date": "2026-05-19", "stem": "In mid-May 2026, an English Football League disciplinary panel threw Southampton out of the season's play-offs after ruling that the club had spied on an opponent's training ahead of their semi-final. Which club's train…", "correct_opt": "Middlesbrough", "model_opt": "Coventry City", "raw": "H, 15", "ok": false, "abstain": false, "conf": 15}]}, {"id": "gpt-5.6-terra", "name": "GPT-5.6 Terra", "provider": "openai", "family": "gpt-5.6", "release": "2026-07-09", "pub_rel": "2026-02", "pub_train": "2026-02", "pub_rel_d": "2026-02-28", "pub_train_d": "2026-02-28", "fit": {"midpoint": "2026-02-12", "width": 32.75, "decay_start": "2025-11-23", "floor_reach": "2026-06-08", "floor": 0.3332, "plateau": 0.9215, "n_items": 1467, "n_rows": 2934, "ci_mid": ["2026-01-20", "2026-03-07"]}, "weekly": [{"d": "2024-07-29", "acc": 0.875, "n": 16, "lo": 0.5291, "hi": 0.9776, "ab": 0.0}, {"d": "2024-08-05", "acc": 0.8846, "n": 26, "lo": 0.6209, "hi": 0.9729, "ab": 0.0}, {"d": "2024-08-12", "acc": 0.8333, "n": 24, "lo": 0.552, "hi": 0.953, "ab": 0.0}, {"d": "2024-08-19", "acc": 0.9231, "n": 26, "lo": 0.6669, "hi": 0.9863, "ab": 0.0}, {"d": "2024-08-26", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2024-09-02", "acc": 0.9062, "n": 32, "lo": 0.6771, "hi": 0.978, "ab": 0.0}, {"d": "2024-09-09", "acc": 0.7692, "n": 26, "lo": 0.4974, "hi": 0.9182, "ab": 0.0}, {"d": "2024-09-16", "acc": 0.8571, "n": 28, "lo": 0.6006, "hi": 0.9599, "ab": 0.0}, {"d": "2024-09-23", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2024-09-30", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2024-10-07", "acc": 0.9286, "n": 14, "lo": 0.5609, "hi": 0.9925, "ab": 0.0}, {"d": "2024-10-14", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2024-10-21", "acc": 0.875, "n": 16, "lo": 0.5291, "hi": 0.9776, "ab": 0.0}, {"d": "2024-10-28", "acc": 0.9167, "n": 24, "lo": 0.6461, "hi": 0.9851, "ab": 0.0}, {"d": "2024-11-04", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2024-11-11", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2024-11-18", "acc": 0.8889, "n": 18, "lo": 0.565, "hi": 0.9801, "ab": 0.0}, {"d": "2024-11-25", "acc": 0.8824, "n": 34, "lo": 0.6566, "hi": 0.9671, "ab": 0.0}, {"d": "2024-12-02", "acc": 0.7273, "n": 22, "lo": 0.4343, "hi": 0.9025, "ab": 0.0}, {"d": "2024-12-09", "acc": 0.9333, "n": 30, "lo": 0.7018, "hi": 0.9881, "ab": 0.0}, {"d": "2024-12-16", "acc": 0.9474, "n": 38, "lo": 0.7536, "hi": 0.9906, "ab": 0.0}, {"d": "2024-12-23", "acc": 0.975, "n": 40, "lo": 0.7995, "hi": 0.9974, "ab": 0.0}, {"d": "2024-12-30", "acc": 1.0, "n": 10, "lo": 0.5655, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-06", "acc": 1.0, "n": 18, "lo": 0.7008, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-13", "acc": 0.9375, "n": 16, "lo": 0.5977, "hi": 0.9934, "ab": 0.0}, {"d": "2025-01-20", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-01-27", "acc": 0.9333, "n": 30, "lo": 0.7018, "hi": 0.9881, "ab": 0.0}, {"d": "2025-02-03", "acc": 0.8929, "n": 28, "lo": 0.6417, "hi": 0.9749, "ab": 0.0}, {"d": "2025-02-10", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-02-17", "acc": 0.9667, "n": 30, "lo": 0.7465, "hi": 0.9965, "ab": 0.0}, {"d": "2025-02-24", "acc": 0.9643, "n": 28, "lo": 0.7324, "hi": 0.9963, "ab": 0.0}, {"d": "2025-03-03", "acc": 0.9375, "n": 32, "lo": 0.7167, "hi": 0.9889, "ab": 0.0}, {"d": "2025-03-10", "acc": 0.9444, "n": 36, "lo": 0.7424, "hi": 0.9901, "ab": 0.0}, {"d": "2025-03-17", "acc": 1.0, "n": 26, "lo": 0.7719, "hi": 1.0, "ab": 0.0}, {"d": "2025-03-24", "acc": 0.95, "n": 20, "lo": 0.6555, "hi": 0.9948, "ab": 0.0}, {"d": "2025-03-31", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-04-07", "acc": 0.9722, "n": 36, "lo": 0.7812, "hi": 0.9971, "ab": 0.0}, {"d": "2025-04-14", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-04-21", "acc": 0.8529, "n": 34, "lo": 0.6225, "hi": 0.9533, "ab": 0.0}, {"d": "2025-04-28", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-05-05", "acc": 0.9286, "n": 28, "lo": 0.6853, "hi": 0.9873, "ab": 0.0}, {"d": "2025-05-12", "acc": 0.8158, "n": 38, "lo": 0.5951, "hi": 0.9303, "ab": 0.0}, {"d": "2025-05-19", "acc": 0.9667, "n": 30, "lo": 0.7465, "hi": 0.9965, "ab": 0.0}, {"d": "2025-05-26", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2025-06-02", "acc": 0.9118, "n": 34, "lo": 0.6924, "hi": 0.9794, "ab": 0.0}, {"d": "2025-06-09", "acc": 1.0, "n": 24, "lo": 0.7575, "hi": 1.0, "ab": 0.0}, {"d": "2025-06-16", "acc": 0.7, "n": 30, "lo": 0.4483, "hi": 0.8701, "ab": 0.0}, {"d": "2025-06-23", "acc": 0.9118, "n": 34, "lo": 0.6924, "hi": 0.9794, "ab": 0.0}, {"d": "2025-06-30", "acc": 0.8421, "n": 38, "lo": 0.6243, "hi": 0.9448, "ab": 0.0}, {"d": "2025-07-07", "acc": 0.875, "n": 24, "lo": 0.5975, "hi": 0.9706, "ab": 0.0}, {"d": "2025-07-14", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2025-07-21", "acc": 0.8667, "n": 30, "lo": 0.6212, "hi": 0.9626, "ab": 0.0}, {"d": "2025-07-28", "acc": 1.0, "n": 30, "lo": 0.7961, "hi": 1.0, "ab": 0.0}, {"d": "2025-08-04", "acc": 1.0, "n": 18, "lo": 0.7008, "hi": 1.0, "ab": 0.0}, {"d": "2025-08-11", "acc": 0.9062, "n": 32, "lo": 0.6771, "hi": 0.978, "ab": 0.0}, {"d": "2025-08-18", "acc": 1.0, "n": 38, "lo": 0.8318, "hi": 1.0, "ab": 0.0}, {"d": "2025-08-25", "acc": 1.0, "n": 42, "lo": 0.8454, "hi": 1.0, "ab": 0.0}, {"d": "2025-09-01", "acc": 1.0, "n": 20, "lo": 0.7225, "hi": 1.0, "ab": 0.0}, {"d": "2025-09-08", "acc": 1.0, "n": 28, "lo": 0.7847, "hi": 1.0, "ab": 0.0}, {"d": "2025-09-15", "acc": 0.9091, "n": 22, "lo": 0.6226, "hi": 0.9838, "ab": 0.0}, {"d": "2025-09-22", "acc": 0.9545, "n": 22, "lo": 0.6785, "hi": 0.9952, "ab": 0.0}, {"d": "2025-09-29", "acc": 0.7727, "n": 22, "lo": 0.4775, "hi": 0.9267, "ab": 0.0}, {"d": "2025-10-06", "acc": 0.9412, "n": 34, "lo": 0.7302, "hi": 0.9895, "ab": 0.0}, {"d": "2025-10-13", "acc": 0.9722, "n": 36, "lo": 0.7812, "hi": 0.9971, "ab": 0.0}, {"d": "2025-10-20", "acc": 0.925, "n": 40, "lo": 0.7306, "hi": 0.9825, "ab": 0.0}, {"d": "2025-10-27", "acc": 0.9286, "n": 42, "lo": 0.7413, "hi": 0.9833, "ab": 0.0}, {"d": "2025-11-03", "acc": 1.0, "n": 26, "lo": 0.7719, "hi": 1.0, "ab": 0.0}, {"d": "2025-11-10", "acc": 0.9643, "n": 28, "lo": 0.7324, "hi": 0.9963, "ab": 0.0}, {"d": "2025-11-17", "acc": 1.0, "n": 22, "lo": 0.7412, "hi": 1.0, "ab": 0.0}, {"d": "2025-11-24", "acc": 0.8, "n": 30, "lo": 0.5481, "hi": 0.9295, "ab": 0.0}, {"d": "2025-12-01", "acc": 0.8, "n": 30, "lo": 0.5481, "hi": 0.9295, "ab": 0.0}, {"d": "2025-12-08", "acc": 0.8182, "n": 22, "lo": 0.523, "hi": 0.9486, "ab": 0.0}, {"d": "2025-12-15", "acc": 0.7895, "n": 38, "lo": 0.5667, "hi": 0.9149, "ab": 0.0}, {"d": "2025-12-22", "acc": 0.7273, "n": 22, "lo": 0.4343, "hi": 0.9025, "ab": 0.0}, {"d": "2025-12-29", "acc": 0.9167, "n": 24, "lo": 0.6461, "hi": 0.9851, "ab": 0.0}, {"d": "2026-01-05", "acc": 0.7308, "n": 26, "lo": 0.4598, "hi": 0.8964, "ab": 0.0}, {"d": "2026-01-12", "acc": 0.6875, "n": 32, "lo": 0.444, "hi": 0.8584, "ab": 0.0}, {"d": "2026-01-19", "acc": 0.7941, "n": 34, "lo": 0.558, "hi": 0.9218, "ab": 0.0}, {"d": "2026-01-26", "acc": 0.6154, "n": 26, "lo": 0.3552, "hi": 0.8229, "ab": 0.0}, {"d": "2026-02-02", "acc": 0.5625, "n": 32, "lo": 0.3318, "hi": 0.769, "ab": 0.0}, {"d": "2026-02-09", "acc": 0.8, "n": 30, "lo": 0.5481, "hi": 0.9295, "ab": 0.0}, {"d": "2026-02-16", "acc": 0.6111, "n": 18, "lo": 0.3092, "hi": 0.8466, "ab": 0.0}, {"d": "2026-02-23", "acc": 0.4643, "n": 28, "lo": 0.2404, "hi": 0.7035, "ab": 0.0}, {"d": "2026-03-02", "acc": 0.5294, "n": 34, "lo": 0.3096, "hi": 0.7384, "ab": 0.0}, {"d": "2026-03-09", "acc": 0.4762, "n": 42, "lo": 0.2834, "hi": 0.6763, "ab": 0.0}, {"d": "2026-03-16", "acc": 0.3214, "n": 28, "lo": 0.1398, "hi": 0.58, "ab": 0.0}, {"d": "2026-03-23", "acc": 0.4, "n": 20, "lo": 0.1682, "hi": 0.6873, "ab": 0.0}, {"d": "2026-03-30", "acc": 0.6562, "n": 32, "lo": 0.4148, "hi": 0.8372, "ab": 0.0}, {"d": "2026-04-06", "acc": 0.5882, "n": 34, "lo": 0.3601, "hi": 0.7839, "ab": 0.0}, {"d": "2026-04-13", "acc": 0.5, "n": 34, "lo": 0.2853, "hi": 0.7147, "ab": 0.0}, {"d": "2026-04-20", "acc": 0.4667, "n": 30, "lo": 0.2481, "hi": 0.6988, "ab": 0.0}, {"d": "2026-04-27", "acc": 0.3158, "n": 38, "lo": 0.1536, "hi": 0.5399, "ab": 0.0}, {"d": "2026-05-04", "acc": 0.3235, "n": 34, "lo": 0.1526, "hi": 0.5595, "ab": 0.0}, {"d": "2026-05-11", "acc": 0.4118, "n": 34, "lo": 0.2161, "hi": 0.6399, "ab": 0.0}, {"d": "2026-05-18", "acc": 0.3846, "n": 26, "lo": 0.1771, "hi": 0.6448, "ab": 0.0}, {"d": "2026-05-25", "acc": 0.3214, "n": 28, "lo": 0.1398, "hi": 0.58, "ab": 0.0}, {"d": "2026-06-01", "acc": 0.4286, "n": 28, "lo": 0.2138, "hi": 0.6741, "ab": 0.0}, {"d": "2026-06-08", "acc": 0.3182, "n": 22, "lo": 0.1236, "hi": 0.6068, "ab": 0.0}, {"d": "2026-06-15", "acc": 0.3846, "n": 26, "lo": 0.1771, "hi": 0.6448, "ab": 0.0}, {"d": "2026-06-22", "acc": 0.3462, "n": 26, "lo": 0.1513, "hi": 0.6112, "ab": 0.0}, {"d": "2026-06-29", "acc": 0.5833, "n": 36, "lo": 0.3614, "hi": 0.776, "ab": 0.0}, {"d": "2026-07-06", "acc": 0.1176, "n": 34, "lo": 0.0329, "hi": 0.3434, "ab": 0.0}, {"d": "2026-07-13", "acc": 0.1842, "n": 38, "lo": 0.0697, "hi": 0.4049, "ab": 0.0}, {"d": "2026-07-20", "acc": 0.3571, "n": 28, "lo": 0.1634, "hi": 0.6124, "ab": 0.0}, {"d": "2026-07-27", "acc": 0.1875, "n": 16, "lo": 0.0445, "hi": 0.5332, "ab": 0.0}], "samples": [{"band": "pre", "date": "2024-08-01", "stem": "In early August 2024, Russia and several Western states carried out the largest exchange of detainees since the Cold War, freeing journalists and dissidents held in Russian jails. Which country acted as the mediator for…", "correct_opt": "Turkey", "model_opt": "Turkey", "raw": "B, 99", "ok": true, "abstain": false, "conf": 99}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, Turkish authorities cut off access to a major social media platform after a government communications official accused its parent company of deleting posts mourning the killing of Hamas political c…", "correct_opt": "Instagram", "model_opt": "Instagram", "raw": "C, 99", "ok": true, "abstain": false, "conf": 99}, {"band": "pre", "date": "2024-08-02", "stem": "In early August 2024, a South American government became the third country in the world to recognise Edmundo González as the rightful winner of Venezuela's disputed presidential election. Which country was it?", "correct_opt": "Argentina", "model_opt": "Ecuador", "raw": "F, 82", "ok": false, "abstain": false, "conf": 82}, {"band": "mid", "date": "2025-11-24", "stem": "In late November 2025, gunmen ambushed a governor's convoy on a road linking a southwestern Yemeni city to the rest of the country, killing about seven people including two of the attackers. Which governorate's governor…", "correct_opt": "Taiz", "model_opt": "Taiz", "raw": "G, 88", "ok": true, "abstain": false, "conf": 88}, {"band": "mid", "date": "2025-11-24", "stem": "In late November 2025, a suicide bomber struck outside the Federal Constabulary headquarters in Peshawar, Pakistan, killing three people before two more gunmen tried to storm the compound. Which militant group claimed t…", "correct_opt": "Jamaat-ul-Ahrar", "model_opt": "Hafiz Gul Bahadur group", "raw": "E, 78", "ok": false, "abstain": false, "conf": 78}, {"band": "mid", "date": "2025-11-24", "stem": "In late November 2025, a U.S. federal judge threw out indictments against James Comey and Letitia James after finding that a prosecutor had been unlawfully installed. Which prosecutor's appointment was ruled invalid?", "correct_opt": "Lindsey Halligan", "model_opt": "Lindsey Halligan", "raw": "D, 95", "ok": true, "abstain": false, "conf": 95}, {"band": "mid", "date": "2025-11-24", "stem": "In late November 2025, seven militants, among them a field commander, were killed in a clash tied to a clan feud in Cotabato in the southern Philippines. Which armed organisation did the slain commander belong to?", "correct_opt": "Moro Islamic Liberation Front", "model_opt": "Bangsamoro Islamic Freedom Fighters", "raw": "E, 65", "ok": false, "abstain": false, "conf": 65}, {"band": "post", "date": "2026-06-08", "stem": "In early June 2026, two basketball players from a well-known Philippine university drowned during a team bonding trip at a coastal resort in Aurora province. Which university did they play for?", "correct_opt": "Ateneo de Manila University", "model_opt": "University of the East", "raw": "C, 58", "ok": false, "abstain": false, "conf": 58}, {"band": "post", "date": "2026-06-08", "stem": "In early June 2026, an Indian state government expelled close to 5,000 Bangladeshi nationals and detained hundreds more in a drive against undocumented migrants launched after its recent state election. Which state carr…", "correct_opt": "West Bengal", "model_opt": "Assam", "raw": "F, 77", "ok": false, "abstain": false, "conf": 77}, {"band": "post", "date": "2026-06-08", "stem": "In early June 2026, eleven people including four police officers died in confrontations between security forces and demonstrators from a banned protest movement in Pakistani-administered Kashmir. Which group's supporter…", "correct_opt": "Joint Awami Action Committee", "model_opt": "Jammu Kashmir Liberation Front", "raw": "G, 78", "ok": false, "abstain": false, "conf": 78}]}];</script> | |
| <script> | |
| (function(){ | |
| 'use strict'; | |
| var M = window.__MODELS__, SKEL = window.__SKELETON__; | |
| var COL = {anthropic:'#eb6834', openai:'#2a78d6'}; | |
| var MEAS='#1baf7a', INK='#0b0b0b', SEC='#52514e', MUT='#898781', GRID='#e1e0d9', BASE='#c3c2b7', SURF='#fcfcfb', STRIPE='#f3f2ee'; | |
| var DAY = 86400000; | |
| function pd(s){ return s ? Date.parse(s + 'T00:00:00Z') : null; } | |
| function iso(t){ return new Date(t).toISOString().slice(0,10); } | |
| function esc(s){ return String(s==null?'':s).replace(/[&<>"]/g, function(c){return {'&':'&','<':'<','>':'>','"':'"'}[c];}); } | |
| var MODES = [ | |
| {key:'cal', label:'Calendar'}, | |
| {key:'rel', label:'Center: release date'}, | |
| {key:'pub', label:'Center: published cutoff'}, | |
| {key:'mid', label:'Center: measured midpoint'} | |
| ]; | |
| function anchor(m, mode){ | |
| if (mode==='rel') return pd(m.release); | |
| if (mode==='pub') return pd(m.pub_rel_d); | |
| if (mode==='mid') return m.fit ? pd(m.fit.midpoint) : null; | |
| return 0; // calendar: absolute | |
| } | |
| function marks(m){ | |
| // green band = midpoint 90% bootstrap CI (where knowledge ends), NOT the | |
| // 80->20 shape width: shallow smeared fits made width-bands absurdly wide. | |
| var f = m.fit || {}; | |
| var mid = pd(f.midpoint), lo = null, hi = null; | |
| if (f.ci_mid && f.ci_mid.length === 2){ lo = pd(f.ci_mid[0]); hi = pd(f.ci_mid[1]); } | |
| if ((lo==null || hi==null) && mid!=null){ | |
| var w = Math.min(21, +f.width || 7)*DAY; // fallback if bootstrap CI missing | |
| lo = mid - w; hi = mid + w; | |
| } | |
| return { | |
| rel: pd(m.release), pubR: pd(m.pub_rel_d), pubT: pd(m.pub_train_d), | |
| t80: lo, t20: hi | |
| }; | |
| } | |
| var mode = 'cal'; | |
| function render(){ | |
| var rows = M.filter(function(m){ return mode==='cal' || anchor(m, mode)!=null; }); | |
| var lo=Infinity, hi=-Infinity; | |
| rows.forEach(function(m){ | |
| var a = anchor(m, mode), k = marks(m); | |
| [k.rel,k.pubR,k.pubT,k.t80,k.t20].forEach(function(t){ | |
| if (t==null) return; var v = t-a; | |
| if (v<lo) lo=v; if (v>hi) hi=v; | |
| }); | |
| }); | |
| if (!isFinite(lo)) { lo = pd('2023-01-01'); hi = pd('2027-01-01'); } | |
| lo -= 45*DAY; hi += 45*DAY; | |
| var labelW=250, W=1160, x0=labelW, x1=W-16, top=58, rowH=32, bot=40; | |
| var H = top + rowH*rows.length + bot; | |
| function X(t){ if(t<lo)t=lo; if(t>hi)t=hi; return x0 + (t-lo)/(hi-lo)*(x1-x0); } | |
| var p = []; | |
| // gridlines | |
| if (mode==='cal'){ | |
| for (var y=2023; y<=2027; y++) for (var q=0;q<4;q++){ | |
| var t = pd(y+'-'+('0'+(q*3+1)).slice(-2)+'-01'); | |
| if (t<lo||t>hi) continue; | |
| var major = q===0, xx=X(t); | |
| p.push('<line x1="'+xx.toFixed(1)+'" y1="'+(top-12)+'" x2="'+xx.toFixed(1)+'" y2="'+(H-bot+8)+'" stroke="'+(major?BASE:GRID)+'" stroke-width="'+(major?1.2:0.7)+'"/>'); | |
| if (major && y<2027) p.push('<text x="'+(xx+4).toFixed(1)+'" y="'+(top-18)+'" fill="'+MUT+'" font-size="12">'+y+'</text>'); | |
| } | |
| } else { | |
| var stepM = (hi-lo) > 900*DAY ? 6 : 3; // month ticks | |
| var mLo = Math.ceil(lo/(30.44*DAY)), mHi = Math.floor(hi/(30.44*DAY)); | |
| for (var mm=mLo; mm<=mHi; mm++){ | |
| if (mm % stepM !== 0) continue; | |
| var t2 = mm*30.44*DAY, xx2 = X(t2), z = mm===0; | |
| p.push('<line x1="'+xx2.toFixed(1)+'" y1="'+(top-12)+'" x2="'+xx2.toFixed(1)+'" y2="'+(H-bot+8)+'" stroke="'+(z?INK:GRID)+'" stroke-width="'+(z?1.6:0.7)+'"/>'); | |
| p.push('<text x="'+(xx2+4).toFixed(1)+'" y="'+(top-18)+'" fill="'+(z?INK:MUT)+'" font-size="11">'+(mm===0?({rel:'release',pub:'published cutoff',mid:'measured midpoint'})[mode]:(mm>0?'+':'')+mm+'m')+'</text>'); | |
| } | |
| } | |
| rows.forEach(function(m,i){ | |
| var a = anchor(m, mode), k = marks(m), yc = top+i*rowH+rowH/2; | |
| var col = COL[(m.provider||'').toLowerCase()] || MUT; | |
| p.push('<g class="mrow" data-id="'+esc(m.id)+'">'); | |
| p.push('<rect x="0" y="'+(yc-rowH/2).toFixed(1)+'" width="'+W+'" height="'+rowH+'" fill="'+(i%2?STRIPE:SURF)+'"/>'); | |
| p.push('<rect x="0" y="'+(yc-6).toFixed(1)+'" width="4" height="12" fill="'+col+'"/>'); | |
| var nm = m.name.length>30 ? m.name.slice(0,29)+'…' : m.name; | |
| p.push('<text x="10" y="'+(yc+4).toFixed(1)+'" fill="'+INK+'" font-size="12">'+esc(nm)+'</text>'); | |
| p.push('<line x1="'+x0+'" y1="'+yc.toFixed(1)+'" x2="'+x1+'" y2="'+yc.toFixed(1)+'" stroke="'+GRID+'" stroke-width="1"/>'); | |
| if (k.pubR!=null && k.pubT!=null && k.pubT>k.pubR) | |
| p.push('<rect x="'+X(k.pubR-a).toFixed(1)+'" y="'+(yc-7).toFixed(1)+'" width="'+(X(k.pubT-a)-X(k.pubR-a)).toFixed(1)+'" height="14" fill="'+col+'" fill-opacity="0.22"/>'); | |
| if (k.pubR!=null) | |
| p.push('<line x1="'+X(k.pubR-a).toFixed(1)+'" y1="'+(yc-9).toFixed(1)+'" x2="'+X(k.pubR-a).toFixed(1)+'" y2="'+(yc+9).toFixed(1)+'" stroke="'+col+'" stroke-width="2"/>'); | |
| if (k.pubT!=null && m.pub_train!==m.pub_rel) | |
| p.push('<line x1="'+X(k.pubT-a).toFixed(1)+'" y1="'+(yc-9).toFixed(1)+'" x2="'+X(k.pubT-a).toFixed(1)+'" y2="'+(yc+9).toFixed(1)+'" stroke="'+col+'" stroke-width="2" stroke-dasharray="2,2"/>'); | |
| if (k.rel!=null){ | |
| var xr = X(k.rel-a); | |
| p.push('<path d="M'+xr.toFixed(1)+','+(yc+3).toFixed(1)+' l4.5,8 h-9 z" fill="'+INK+'"/>'); | |
| } | |
| if (k.t80!=null && k.t20!=null) | |
| p.push('<rect x="'+X(k.t80-a).toFixed(1)+'" y="'+(yc-4).toFixed(1)+'" width="'+Math.max(2, X(k.t20-a)-X(k.t80-a)).toFixed(1)+'" height="8" rx="3" fill="'+MEAS+'" fill-opacity="0.8"/>'); | |
| else if (SKEL) | |
| p.push('<text x="'+x1+'" y="'+(yc+4).toFixed(1)+'" fill="'+MUT+'" font-size="10" text-anchor="end">published only</text>'); | |
| p.push('</g>'); | |
| }); | |
| var svg = document.getElementById('tl'); | |
| svg.setAttribute('viewBox', '0 0 '+W+' '+H); | |
| svg.innerHTML = p.join(''); | |
| wire(); | |
| } | |
| // ---------- tooltips + click ---------- | |
| function tipHtml(m){ | |
| var L = ['<b>'+esc(m.id)+'</b>']; | |
| if (m.name && m.name!==m.id) L.push(esc(m.name)); | |
| L.push('provider: '+esc(m.provider)+' · family: '+esc(m.family)); | |
| L.push('release: '+esc(m.release)); | |
| L.push('published reliable: '+esc(m.pub_rel)+' · training: '+esc(m.pub_train)); | |
| if (m.fit){ | |
| L.push('measured midpoint: '+esc(m.fit.midpoint)+' · width '+esc(m.fit.width)+'d'); | |
| var ds = pd(m.fit.decay_start), pr = pd(m.pub_rel_d); | |
| if (ds!=null && pr!=null){ | |
| var dd = Math.round((ds-pr)/DAY); | |
| L.push('decay starts '+esc(m.fit.decay_start)+' ('+Math.abs(dd)+'d '+(dd>0?'after':'before')+' published cutoff)'); | |
| } | |
| L.push('plateau '+esc(m.fit.plateau)+' · floor '+esc(m.fit.floor)); | |
| L.push('<i>click for the weekly series + sample answers</i>'); | |
| } | |
| return L.join('<br>'); | |
| } | |
| function wire(){ | |
| var tip = document.getElementById('tip'); | |
| Array.prototype.forEach.call(document.querySelectorAll('.mrow'), function(g){ | |
| var m = M.filter(function(x){return x.id===g.getAttribute('data-id');})[0]; | |
| if (!m) return; | |
| g.addEventListener('mousemove', function(e){ | |
| tip.innerHTML = tipHtml(m); tip.style.display='block'; | |
| var x=e.clientX+14, y=e.clientY+14; | |
| if (x+400>window.innerWidth) x=Math.max(4,e.clientX-400); | |
| if (y+190>window.innerHeight) y=Math.max(4,e.clientY-190); | |
| tip.style.left=x+'px'; tip.style.top=y+'px'; | |
| }); | |
| g.addEventListener('mouseleave', function(){ tip.style.display='none'; }); | |
| g.addEventListener('click', function(){ openDetail(m); }); | |
| }); | |
| } | |
| // ---------- detail panel ---------- | |
| function fitAcc(f, t){ | |
| var mid = pd(f.midpoint), w = (+f.width||1)*DAY; | |
| return (+f.floor) + ((+f.plateau)-(+f.floor)) / (1 + Math.exp((t-mid)/w)); | |
| } | |
| function openDetail(m){ | |
| var el = document.getElementById('detail'); | |
| if (!m.fit){ el.style.display='none'; return; } | |
| var f = m.fit, wk = m.weekly||[], col = COL[(m.provider||'').toLowerCase()]||MUT; | |
| var W=1120, H=332, L=46, R=14, T=50, B=34; | |
| var HALO = ' stroke="'+SURF+'" stroke-width="3.5" paint-order="stroke" stroke-linejoin="round"'; | |
| var lo = pd('2024-07-15'), hi = pd('2026-08-15'); | |
| function X(t){ return L + (t-lo)/(hi-lo)*(W-L-R); } | |
| function Y(v){ return T + (1-v)*(H-T-B); } | |
| var p=[]; | |
| [0,0.25,0.5,0.75,1].forEach(function(v){ | |
| p.push('<line x1="'+L+'" y1="'+Y(v).toFixed(1)+'" x2="'+(W-R)+'" y2="'+Y(v).toFixed(1)+'" stroke="'+GRID+'" stroke-width="0.7"/>'); | |
| p.push('<text x="'+(L-6)+'" y="'+(Y(v)+4).toFixed(1)+'" fill="'+MUT+'" font-size="10" text-anchor="end">'+(v*100)+'%</text>'); | |
| }); | |
| for (var y=2024; y<=2026; y++) for (var q=0;q<4;q++){ | |
| var t = pd(y+'-'+('0'+(q*3+1)).slice(-2)+'-01'); | |
| if (t<lo||t>hi) continue; | |
| p.push('<line x1="'+X(t).toFixed(1)+'" y1="'+T+'" x2="'+X(t).toFixed(1)+'" y2="'+(H-B)+'" stroke="'+GRID+'" stroke-width="0.6"/>'); | |
| p.push('<text x="'+(X(t)+3).toFixed(1)+'" y="'+(H-B+14)+'" fill="'+MUT+'" font-size="10">'+y+'-'+('0'+(q*3+1)).slice(-2)+'</text>'); | |
| } | |
| // reference lines: floor & plateau | |
| [['floor', f.floor], ['plateau', f.plateau]].forEach(function(fp){ | |
| if (fp[1]==null) return; | |
| p.push('<line x1="'+L+'" y1="'+Y(+fp[1]).toFixed(1)+'" x2="'+(W-R)+'" y2="'+Y(+fp[1]).toFixed(1)+'" stroke="'+MUT+'" stroke-width="1" stroke-dasharray="1,4"/>'); | |
| p.push('<text x="'+(W-R)+'" y="'+(Y(+fp[1])-3).toFixed(1)+'" fill="'+MUT+'" font-size="10" text-anchor="end"'+HALO+'>'+fp[0]+' '+(+fp[1]).toFixed(2)+'</text>'); | |
| }); | |
| // markers: published cutoffs, release, midpoint CI. Labels are collected and | |
| // laid out afterwards with lane-based collision avoidance (stagger + edge flip). | |
| var vlabels = []; | |
| function vline(t, color, dash, label){ | |
| if (t==null || t<lo || t>hi) return; | |
| p.push('<line x1="'+X(t).toFixed(1)+'" y1="'+T+'" x2="'+X(t).toFixed(1)+'" y2="'+(H-B)+'" stroke="'+color+'" stroke-width="1.6"'+(dash?' stroke-dasharray="4,3"':'')+'/>'); | |
| if (label) vlabels.push({x: X(t), text: label, color: color}); | |
| } | |
| function layoutVLabels(){ | |
| vlabels.sort(function(a,b){ return a.x-b.x; }); | |
| var laneEnd = [-1e9,-1e9,-1e9,-1e9]; | |
| vlabels.forEach(function(lb){ | |
| var w = lb.text.length*5.7 + 8; | |
| var anchor = 'start', xs = lb.x+4, xe = xs+w; | |
| if (xe > W-R-60){ anchor='end'; xs = lb.x-4-w; xe = lb.x-4; } | |
| var lane = 0; | |
| while (lane < 3 && xs < laneEnd[lane]+8) lane++; | |
| laneEnd[lane] = xe; | |
| var ly = 12 + lane*13; // labels live in the header strip above the plot | |
| p.push('<line x1="'+lb.x.toFixed(1)+'" y1="'+(ly+2)+'" x2="'+lb.x.toFixed(1)+'" y2="'+T+'" stroke="'+lb.color+'" stroke-width="1" opacity="0.5"/>'); | |
| p.push('<text x="'+(anchor==='start'?lb.x+4:lb.x-4).toFixed(1)+'" y="'+ly+'" fill="'+lb.color+'" font-size="10" text-anchor="'+anchor+'"'+HALO+'>'+esc(lb.text)+'</text>'); | |
| }); | |
| } | |
| if (f.ci_mid && f.ci_mid.length===2){ | |
| var c0=pd(f.ci_mid[0]), c1=pd(f.ci_mid[1]); | |
| if (c0!=null&&c1!=null) p.push('<rect x="'+X(c0).toFixed(1)+'" y="'+T+'" width="'+(X(c1)-X(c0)).toFixed(1)+'" height="'+(H-T-B)+'" fill="'+MEAS+'" fill-opacity="0.10"/>'); | |
| } | |
| vline(pd(m.pub_rel_d), col, false, 'published reliable '+m.pub_rel); | |
| if (m.pub_train!==m.pub_rel) vline(pd(m.pub_train_d), col, true, 'training '+m.pub_train); | |
| vline(pd(m.release), INK, true, 'release'); | |
| vline(pd(f.midpoint), MEAS, false, 'measured midpoint '+f.midpoint); | |
| // fitted curve | |
| var line=[]; | |
| for (var t=lo; t<=hi; t+=7*DAY) line.push((line.length?'L':'M')+X(t).toFixed(1)+','+Y(Math.max(0,Math.min(1,fitAcc(f,t)))).toFixed(1)); | |
| p.push('<path d="'+line.join('')+'" fill="none" stroke="'+MEAS+'" stroke-width="2.2" opacity="0.9"/>'); | |
| // weekly points + wilson whiskers | |
| wk.forEach(function(s){ | |
| var t = pd(s.d); if (t==null||s.acc==null) return; | |
| if (s.lo!=null && s.hi!=null) | |
| p.push('<line x1="'+X(t).toFixed(1)+'" y1="'+Y(s.lo).toFixed(1)+'" x2="'+X(t).toFixed(1)+'" y2="'+Y(s.hi).toFixed(1)+'" stroke="'+col+'" stroke-width="1" opacity="0.28"/>'); | |
| p.push('<circle cx="'+X(t).toFixed(1)+'" cy="'+Y(s.acc).toFixed(1)+'" r="2.6" fill="'+col+'" opacity="0.85"><title>'+esc(s.d)+' · acc '+(s.acc*100).toFixed(0)+'% · n='+s.n+(s.ab?(' · abstain '+(s.ab*100).toFixed(0)+'%'):'')+'</title></circle>'); | |
| }); | |
| layoutVLabels(); | |
| var ciTxt = (f.ci_mid&&f.ci_mid.length===2) ? (' (90% CI '+f.ci_mid[0]+' … '+f.ci_mid[1]+')') : ''; | |
| var html = '<button class="close" id="dclose">close ×</button>' | |
| + '<h3>'+esc(m.name)+' <span style="color:'+SEC+';font-weight:400">('+esc(m.id)+')</span></h3>' | |
| + '<p class="meta">release '+esc(m.release)+' · published reliable <b>'+esc(m.pub_rel)+'</b>' | |
| + (m.pub_train!==m.pub_rel ? ' · training <b>'+esc(m.pub_train)+'</b>' : '') | |
| + ' | measured midpoint <b>'+esc(f.midpoint)+'</b>'+esc(ciTxt)+' · width '+esc(f.width)+'d · decay starts '+esc(f.decay_start)+' · floor reached '+esc(f.floor_reach) | |
| + '<br>plateau '+(+f.plateau).toFixed(3)+' · floor '+(+f.floor).toFixed(3)+' · '+esc(f.n_items)+' filtered items · '+esc(f.n_rows)+' answers (2 shuffled permutations, minimal reasoning)' | |
| + '<br><span style="color:'+MUT+'">series: weekly accuracy on the headline set (detail+categorical items, grounded, not guessable by the 12 pre-window control models: pooled control mean <0.5 and <2 controls solving both permutations); whiskers = 95% Wilson intervals; curve = fitted logistic floor+(plateau−floor)·σ((midpoint−t)/width).</span></p>' | |
| + '<svg viewBox="0 0 '+W+' '+H+'">'+p.join('')+'</svg>'; | |
| var sm = m.samples||[]; | |
| if (sm.length){ | |
| var tag = {pre:'pre-cutoff', mid:'transition', post:'post-cutoff'}; | |
| html += '<table><tr><th style="width:86px">event</th><th style="width:84px"></th><th>question</th><th style="width:26%">model answered</th><th style="width:22%">correct answer</th><th style="width:56px">conf</th></tr>'; | |
| sm.forEach(function(s){ | |
| var ans = s.abstain ? '<i>(abstained)</i> “'+esc(s.raw)+'”' : esc(s.model_opt==null?s.raw:s.model_opt); | |
| html += '<tr><td>'+esc(s.date)+'<br><span class="bandtag">'+tag[s.band]+'</span></td>' | |
| + '<td class="'+(s.ok?'ok':'bad')+'">'+(s.ok?'✓ right':'✗ wrong')+'</td>' | |
| + '<td>'+esc(s.stem)+'</td><td>'+ans+'</td><td>'+esc(s.correct_opt)+'</td>' | |
| + '<td>'+(s.conf==null?'—':esc(s.conf))+'</td></tr>'; | |
| }); | |
| html += '</table>'; | |
| } | |
| el.innerHTML = html; el.style.display='block'; | |
| document.getElementById('dclose').addEventListener('click', function(){ el.style.display='none'; }); | |
| el.scrollIntoView({behavior:'smooth', block:'nearest'}); | |
| } | |
| document.addEventListener('keydown', function(e){ if (e.key==='Escape') document.getElementById('detail').style.display='none'; }); | |
| // ---------- mode buttons ---------- | |
| var bar = document.getElementById('modes'); | |
| MODES.forEach(function(md){ | |
| var b = document.createElement('button'); | |
| b.textContent = md.label; b.setAttribute('data-k', md.key); | |
| if (md.key===mode) b.className='on'; | |
| b.addEventListener('click', function(){ | |
| mode = md.key; | |
| Array.prototype.forEach.call(bar.querySelectorAll('button'), function(x){ x.className = x.getAttribute('data-k')===mode?'on':''; }); | |
| render(); | |
| }); | |
| bar.appendChild(b); | |
| }); | |
| render(); | |
| })(); | |
| </script></body></html> |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment