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.

hackermondev

hi, i'm daniel. i'm a 15-year-old with some programming experience and i do a little bug hunting in my free time. here's the insane story of how I found a single bug that affected over half of all Fortune 500 companies:

say hello to zendesk

If you've spent some time online, you’ve probably come across Zendesk.

Zendesk is a customer service tool used by some of the world’s top companies. It’s easy to set up: you link it to your company’s support email (like support@company.com), and Zendesk starts managing incoming emails and creating tickets. You can handle these tickets yourself or have a support team do it for you. Zendesk is a billion-dollar company, trusted by big names like Cloudflare.

Personally, I’ve always found it surprising that these massive companies, worth billions, rely on third-party tools like Zendesk instead of building their own in-house ticketing systems.

your weakest link

andreis

print("\n".join("".join("*" if i in (0, 9, j, 9-j) else " " for j in range(10)) for i in range(10)))
print("====")
print("".join(map(lambda i: ("*" if i//10 in (0, 9, i%10, 9-(i%10)) else " ") + ("\n" if (i+1) % 10 == 0 else ""), range(100))))
print("====")
print("".join(map(lambda i: ([""]*9+["\n"])[(i>>1)%10] if i&1 else "*" if (i>>1)//10 in (0, 9, (i>>1)%10, 9-((i>>1)%10)) else " ", range(200))))

cpursley

PostgreSQL is Enough

See the new site: https://postgresisenough.dev

---

• Simplify: move code into database functions
• Just Use Postgres for Everything
• Just use Postgres
• PostgreSQL is the worlds’ best database

postmalloc

// A handy bookmarklet to display comments from the top-rated Hacker News thread related to the current page
// Written with the help of GPT-4

javascript:(function() {
  const createCommentElement = (comment, depth) => {
    const commentWrapper = document.createElement('div');
    commentWrapper.style.paddingLeft = (depth * 20) + 'px';
    commentWrapper.style.marginBottom = '10px';
    commentWrapper.style.marginLeft = '10px';
    commentWrapper.style.color = '#333';

rain-1

This worked on 14/May/23. The instructions will probably require updating in the future.

llama is a text prediction model similar to GPT-2, and the version of GPT-3 that has not been fine tuned yet. It is also possible to run fine tuned versions (like alpaca or vicuna with this. I think. Those versions are more focused on answering questions)

Note: I have been told that this does not support multiple GPUs. It can only use a single GPU.

It is possible to run LLama 13B with a 6GB graphics card now! (e.g. a RTX 2060). Thanks to the amazing work involved in llama.cpp. The latest change is CUDA/cuBLAS which allows you pick an arbitrary number of the transformer layers to be run on the GPU. This is perfect for low VRAM.

• Clone llama.cpp from git, I am on commit 08737ef720f0510c7ec2aa84d7f70c691073c35d.

kconner

macOS Internals

Understand your Mac and iPhone more deeply by tracing the evolution of Mac OS X from prelease to Swift. John Siracusa delivers the details.

Starting Points

How to use this gist

You've got two main options:

andreis

# turn until you get three hexagons changed to get here
H1 = ["AC", "AB", "ACD"]
# turn until you get two hexagons changed to get here
H2 = ["BCF", "ACF", "BF"]
# turn until you get three hexagons changed to get here
H3 = ["CD", "BC", "AB", "AF", "EF", "CDE"]

# this is my status, look at the hexagons to determine which ones are open
status = set("ABDE") # 110110

cecilemuller

name: Test

on:
  push:
    branches:
    - main
    - features/**
    - dependabot/**
  pull_request:
    branches:

mGBUfLn9

I was drawn to programming, science, technology and science fiction
ever since I was a little kid. I can't say it's because I wanted to
make the world a better place. Not really. I was simply drawn to it
because I was drawn to it. Writing programs was fun. Figuring out how
nature works was fascinating. Science fiction felt like a grand
adventure.

Then I started a software company and poured every ounce of energy
into it. It failed. That hurt, but that part is ok. I made a lot of
mistakes and learned from them. This experience made me much, much