2025 ICML ICML 2025

RE-Bench: Evaluating Frontier AI R&D Capabilities of Language Model Agents against Human Experts

👥 Mega-Team — 22 authors