| #!/bin/bash | |
| # Show real memory usage (working set) per container | |
| # working set = memory.current - inactive_file (excludes reclaimable page cache) | |
| printf "%-30s %10s %10s %10s\n" "CONTAINER" "TOTAL" "CACHE" "WORKING SET" | |
| printf "%-30s %10s %10s %10s\n" "---------" "-----" "-----" "-----------" | |
| docker ps --format '{{.ID}} {{.Names}}' | while read id name; do | |
| cgroup=$(docker inspect --format '{{.Id}}' "$id") |
dv serve adds an HTTP API server to dv, exposing the full CLI surface over
JSON + Server-Sent Events. The primary consumer is Jarvis (running in a Docker
container), which reaches the host via host.docker.internal or the Docker
bridge gateway. Any HTTP client works — curl, scripts, future web UIs.
Design principles:
| #!/usr/bin/env ruby | |
| # test_fork_crash.rb | |
| # | |
| # Demonstrates mini_racer fork safety — specifically that using an inherited | |
| # MiniRacer::Context across a fork boundary WITHOUT single_threaded mode causes | |
| # a deadlock/hang every time. | |
| # | |
| # Safety mechanism: each scenario runs as an isolated subprocess wrapped in an | |
| # OS-level `timeout`, so a hang can never lock up this harness. | |
| # |
Papers from arXiv, last 5 days. Full 90-day database populating overnight.
Paper: A Decision-Theoretic Formalisation of Steganography With Applications to LLM Monitoring
Published: 26 Feb 2026 · cs.AI
Paper: arXiv:2602.23163
Authors: Anwar, Piskorz, Baek, Africa, Weatherall, Tegmark, de Witt, van der Schaar, Krueger
Venue: ICML 2026
Paper: arXiv:2602.23163
Authors: Anwar, Piskorz, Baek, Africa, Weatherall, Tegmark, de Witt, van der Schaar, Krueger
Venue: ICML 2026
1,600 papers. Here's what actually matters.
This week's dominant theme: the gap between what scaling should fix and what it actually fixes. A standout paper argues that VLMs can't reason about visual scenes not because they're undertrained, but because humans never describe what's obvious — a structural problem no amount of compute solves. Meanwhile, medical AI took a meaningful step forward with RL-based free-form reasoning (not just multiple-choice), and the efficiency crowd got a win with memory-optimised training that makes 7B+ models accessible without 100GB of VRAM. The biosecurity community got a sobering data point: LLMs do meaningfully uplift novices on dual-use biology tasks.
This week's standout theme: limits. Multiple papers poked holes in the "scale fixes everything" assumption — from VLMs that fail at basic reasoning because their training data is systematically biased, to memory constraints that make state-of-the-art tree search impractical. On the application side, medical AI made a real leap with open-ended clinical reasoning (not just multiple-choice), and a sobering biosecurity study measured whether LLMs actually help novices do dangerous things. Spoiler: the answer isn't clean. With 1,600 papers across five days, the field's output velocity continues to be staggering — and the signal-to-noise problem is getting worse.
1,600 papers indexed. Here's what actually matters.
This week's dominant theme: the scaling gospel taking hits from multiple directions. A notable paper argues that VLMs fail at visual reasoning not because they're too small, but because human language itself systematically omits the visual facts models need to learn — a problem more training data won't fix. Meanwhile, medical AI pushed past the multiple-choice crutch with RL-based free-form reasoning, and 3D reconstruction got a fix for the quadratic memory scaling wall that's been quietly killing the field. No moonshot drops this week, but a lot of quiet assumptions got kicked.
3,200 papers indexed. Here's what actually matters.
Late February was dominated by two themes pulling in opposite directions: models getting more capable in narrow domains (medical reasoning, omni-modal, agentic coding), while researchers kept poking holes in the assumptions underpinning those gains. The most interesting result of the period isn't a new benchmark SOTA — it's the finding that VLM reasoning failures aren't a scale problem, they're a data bias problem baked into how humans caption images. Meanwhile, the biosecurity community got a properly rigorous study on LLM uplift for dual-use biology tasks, and the results are uncomfortable. On the infrastructure side, FlashOptim quietly solved a problem that's been annoying everyone training models without enterprise-grade hardware.