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.

realvjy

// Lines
float hash( float n ) {
    return fract(sin(n)*753.5453123);
}

// Slight modification of iq's noise function.
float noise(vector_float2 x )
{
    vector_float2 p = floor(x);
    vector_float2 f = fract(x);

amosgyamfi

//
//  AirPodsMaxAnimation.swift
//  OpenvisionOS
//
//  Created by Amos Gyamfi on 12.11.2023.
//

import SwiftUI
import RealityKit
import RealityKitContent

alexwidua

import SwiftUI

// 1. Use looped H/VStacks to create a grid 
// 2. Conditionally increase spacing to grow/shrink the grid
// 3. Calculate the distance of each dot to the center and use the value to stagger the animation 
//4. Add random delay on top of the staggered delay value

struct ContentView: View {
    
    // const & state