KS Chan mrkschan

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.

Short link to this page: http://caseywatts.com/ptt

Other gists & tricks: http://caseywatts.com/gists-and-tricks

Unrelated update: my book is out! Debugging Your Brain is an applied psychology / self-help book

Push To Talk in a Google Hangout (Meet)

Save this bookmarklet. Right-click on boomarks toolbar Add Page...

Name: PTT (push to talk) or whatever you'd like (maybe short so it stays on your bookmarks toolbar)

	"""
	The most atomic way to train and run inference for a GPT in pure, dependency-free Python.
	This file is the complete algorithm.
	Everything else is just efficiency.

	@karpathy
	"""

	import os # os.path.exists
	import math # math.log, math.exp

	#!/usr/bin/python3

	# This is used for calculating of birthday paradox for large values
	# We're using approximation explained here: http://preshing.com/20110504/hash-collision-probabilities/
	# Beware that approximation isn't very precise at smaller params N and K, but gets more precise with large numbers

	# see https://docs.python.org/3/library/decimal.html#module-decimal
	from decimal import *

	# setting decimal precision

	'''
	Modified from https://github.com/wuyunfeng/Python-FastCGI-Client
	'''

	from urlparse import urlparse as parse_url
	import random
	import socket
	import sys

	FROM debian:stretch

	RUN set -ex; \
	apt-get update; \
	apt-get install -y --no-install-recommends \
	build-essential \
	curl \
	python-pip python-setuptools python-virtualenv \
	; \
	rm -rf /var/lib/apt/lists/*

	package main

	import (
	"fmt"
	"sync"
	"time"
	)

	type TimedLock struct {
	mutex chan (uint)

	import http.server
	import socketserver

	PORT = 8080

	class H(http.server.SimpleHTTPRequestHandler):
	def do_GET(self):
	self.send_response(200)
	self.end_headers()