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.

jackdomleo7

/*! NOTE: These are just recommended default global styles, edit as required */

@import ('Import reset stylesheet here, (I recommend modern-normalize) or even better, import the reset stylesheet in the HTML as the first imported stylesheet');

::selection { /* Optional */
  /* It can be really hard to read highlighted text with a text-shadow, it should be removed when selected */
  text-shadow: none;
  
  /* NOTE: Using this means the color and background-color are set to transparent for selected text... */
  /* So you can customise your styles below */

madcodemonkey

using ServiceReference1;
using System;
using System.Collections.Generic;
using System.IO;
using System.ServiceModel;
using System.Threading.Tasks;

namespace SSRSSimple
{
    class Program

roblabs

#!/bin/bash -e

# --------------------------------------------------------
# Generate iOS Launch items from a single image
# (c) 2017 ePi Rational, Inc.
#
# Modified from original script which generated app icons for iOS,
# originally written by
# Ben Clayton, Calvium Ltd.
#  https://gist.github.com/benvium/2be6d673aa9ac284bb8a

samgiles

// [B](f: (A) ⇒ [B]): [B]  ; Although the types in the arrays aren't strict (:
Array.prototype.flatMap = function(lambda) { 
    return Array.prototype.concat.apply([], this.map(lambda)); 
};