Chen Bo Calvin Zhang

Reliable Weak-to-Strong Monitoring of LLM Agents

MoReBench: Evaluating Procedural and Pluralistic Moral Reasoning in Language Models, More than Outcomes

ResearchRubrics: A Benchmark of Prompts and Rubrics For Evaluating Deep Research Agents

Beyond Seeing: Evaluating Multimodal LLMs on Tool-Enabled Image Perception, Transformation, and Reasoning

SWE-Bench Pro: Can AI Agents Solve Long-Horizon Software Engineering Tasks?

TutorBench: A Benchmark To Assess Tutoring Capabilities Of Large Language Models

ORSO: Accelerating Reward Design via Online Reward Selection and Policy Optimization