Research paper
High importance
@karpathy
Importance score: 3 • Posted: February 27, 2026 at 23:08
Score
3
I had the same thought so I've been playing with it in nanochat. E.g. here's 8 agents (4 claude, 4 codex), with 1 GPU each running nanochat experiments (trying to delete logit softcap without regression). The TLDR is that it doesn't work and it's a mess... but it's still very pretty to look at :) ...
How come the NanoGPT speedrun challenge is not fully AI automated research by now?
Media
Grok reasoning
Andrej Karpathy demos multi-agent AI research organization attempting automated ML experiments, highlighting current limitations and future potential in AI-driven research.
Likes
6,560
Reposts
569
Views
994,860
Tweet ID: 2027521323275325622
Prompt source: ai-news
Fetched at: February 28, 2026 at 19:53