Benchmarking LLMs at the Game Of Science (Eleusis)

Evaluates iterative hypothesis testing rather than fixed-answer recall, the same loop coding agents use when debugging, so results speak to a failure mode engineers hit daily.

• benchmark — A standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.

• videoTeaching Tiny Models to Prove Hard Theorems | Lewis Tunstall | HF ML Club India EP1
• videoRoPE: Understanding Rotary Positional Embeddings in transformers
• videoHugging Face Journal Club: AI Research Preference Models
• videoMoE Token Routing Explained: How Mixture of Experts Works (with Code)

• articleGames as Model Eval: 1-Click Deploy AI Town on Fly.iofly.io
• postExplorationBench tests how models explore, with exactly verifiable gradingTencent Hy
• videoThe Good, the Bad, and the Ugly: Why Coding Benchmarks Are BrokenAI Engineer

Checking sign-in…

Loading comments…