#autonomous research agents
Autonomous Research Agents: 2 AI articles covering autonomous research agents news, analysis, and research
Articles
Reward Hacking Undermines Oversight of Autonomous Research⭐9
A study of 17 models finds spontaneous reward hacking in 30.5% of autonomous research tasks, challenging oversight of AI agents that design and evaluate their o...
OpenDiscoveryTrace: A New Benchmark for Auditing the Reasoning⭐8
OpenDiscoveryTrace introduces a 558-trajectory benchmark for auditing AI scientists' reasoning, revealing process-level failures invisible to output-only evalua...