Back to article

karpathy/autoresearch: AI agents running research on single-GPU nanochat training automatically (opens in new tab)

Covered by 30 sources including DEV Community, datadoghq.comDiscussed on Hacker News, Lobsters, and DEV

Covered in 35 articles

DEV Community·

Loopcraft: Stop Prompting, Start Designing Loops

Discussed on DEV

DEV Community·

Open Source Project of the Day (#83): Darwin Skill - A Karpathy-Inspired 'Ratchet' System for Infinite AI Skill Evolution

Discussed on DEV

DEV Community·

Fabricate, Collaborate, Elaborate, Delegate, Validate: Why My Bootstrapping Command Produces Better One-Shot Code

Discussed on DEV

DEV Community·

An Evolving Strategy for Knowledge Work: From Human-In-the-Loop to Human-Before-the-Loop

Discussed on DEV

datadoghq.com·

How we made a SQL query optimization agent 59% more accurate using autoresearch and LLM Observability

Zed Industries·

The Community Champions Program

huggingface.co·

Karpathy's autoresearch, 50 DPO experiments, 300 human judges

Discussed on Hacker News

Interesting Engineering++·

Designing Loops - A Practitioner's Short Field Guide

Discussed on Substack

I Let an AI Agent Run 40 Experiments While I Slept

RSI is the new AGI — and it’s just as hard to pin down

In other languages

Andrej Karpathy, Anthropic에 합류

Как мы строили eval и автоматизировали оптимизацию AI-агента для умного помощника

Как правильно заLOOPить Fable 5