#nanochat

Andrej Karpathy has published autoresearch, a minimal repo that lets AI agents iterate on a stripped-down nanochat training loop overnight. The project turns agent evaluation into a closed-loop research workflow with fixed 5-minute runs, Git branches, and validation-loss-based selection.

A Hacker News submission highlighted Andrej Karpathy's Autoresearch repo, a minimal setup where an AI agent edits one training file, runs fixed 5-minute experiments, and keeps only changes that improve `val_bpb`.