karpathy

LLM Wiki

A pattern for building personal knowledge bases using LLMs.

This is an idea file, it is designed to be copy pasted to your own LLM Agent (e.g. OpenAI Codex, Claude Code, OpenCode / Pi, or etc.). Its goal is to communicate the high level idea, but your agent will build out the specifics in collaboration with you.

The core idea

Most people's experience with LLMs and documents looks like RAG: you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works, but the LLM is rediscovering knowledge from scratch on every question. There's no accumulation. Ask a subtle question that requires synthesizing five documents, and the LLM has to find and piece together the relevant fragments every time. Nothing is built up. NotebookLM, ChatGPT file uploads, and most RAG systems work this way.

tiptopcoder

import { NextPageContext } from "next";

const Error = ({ statusCode }) => {
  return (
    <p>
      {statusCode
        ? `An error ${statusCode} occurred on server`
        : "An error occurred on client"}
    </p>
  );

mavieth

[
  "aol.com",
  "att.net",
  "comcast.net",
  "facebook.com",
  "gmail.com",
  "gmx.com",
  "googlemail.com",
  "google.com",
  "hotmail.com",

xavierfoucrier

GPG signing –

Here is a short guide that will help you setup your environment to create signed commits or signed tags with Git locally. This has been extensively tested on Windows with Git and the Github Desktop application: I use it every day for my professional development projects.

I you face any issue, feel free to leave a comment below.

Summary

1. Sign commits or tags
2. Key passphrase
3. Disable signatures
4. Renew a GPG key

bpesquet

* text=auto

# Unity files
*.meta -text -merge=unityamlmerge
*.unity -text -merge=unityamlmerge
*.asset -text -merge=unityamlmerge
*.prefab -text -merge=unityamlmerge

# Image formats
*.psd filter=lfs diff=lfs merge=lfs -text

wchargin

digraph images_demo_gaussian_blur {
    rankdir = "BT";
    labelloc = "b";
    fontname = "Roboto";
    node [fontname = "Roboto"];
    edge [fontname = "Roboto"];

    gr_decode [label="decode_image", shape=box, style=rounded];

    gr_image [label="image", shape=box, style=rounded];

trusktr

blah