The Automated LLM Speedrunning Benchmark measures how well LLM agents can reproduce previous innovations and discover new ones in language modeling. - View it on GitHub
Star
109
Rank
262277