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.

jdrew1303

Introduction

The computer driven markets for instruments like stocks and exchange traded stock options, have transformed finance and the flow of capital. These markets are enabled by order matching engines (and the infrastructure that supports this software). Before computer trading networks and matching engines, stocks where traded on cavernous exchange floors and transaction costs where high. When electronic trading fully matured, floor traders were a fading anachronism and transaction costs had been reduced to pennies a share in many cases. Electronic trading could not exist without advanced network infrastructure, but without the software matching engines no shares would change hands. The computer trading networks, the matching engine software has also created a concentrated nexus of potential failure. Failures in these systems have increased as the frequency and volume on the electronic networks has increased. The position of order matching engines in the trading infrastructure makes these systems o

culmat

package common;
/**
 * <blockquote>
 * <pre>
 * Map<String, Integer> myMap = map("a", 1).with("b", 2);
 * </pre>
 * </blockquote>
 * 
 * The first tuple determines the map type, so cast as appropriate: 
 * <blockquote>

Lewiscowles1986

#!/bin/bash

if [ "$EUID" -ne 0 ]
	then echo "Must be root"
	exit
fi

ADAPTER="eth0"

# Allow overriding from eth0 by passing in a single argument

Lewiscowles1986

#!/bin/bash
#
# This version uses September 2017 august stretch image, please use this image
#

if [ "$EUID" -ne 0 ]
	then echo "Must be root"
	exit
fi

gtallen1187

"A little bit of slope makes up for a lot of y-intercept"

01/13/2012. From a lecture by Professor John Ousterhout at Stanford, class CS140

Here's today's thought for the weekend. A little bit of slope makes up for a lot of Y-intercept.

[Laughter]

tsenart

package main

import (
	"net/http"
	"database/sql"
	"fmt"
	"log"
	"os"
)

asbestos

class StreamingStatementCreator implements PreparedStatementCreator {
    private final String sql;

    public StreamingStatementCreator(String sql) {
        this.sql = sql;
    }

    @Override
    public PreparedStatement createPreparedStatement(Connection connection) throws SQLException {
        final PreparedStatement statement = connection.prepareStatement(sql, ResultSet.TYPE_FORWARD_ONLY, ResultSet.CONCUR_READ_ONLY);

kevin-smets

Default

Powerlevel10k

chrisdarroch

#!/bin/sh
 
# check for where the latest version of IDEA is installed
IDEA=`ls -1d /Applications/IntelliJ\ * | tail -n1`
wd=`pwd`

# were we given a directory?
if [ -d "$1" ]; then
#  echo "checking for things in the working dir given"
  wd=`ls -1d "$1" | head -n1`