Kun Wan

Vision-Zero: Scalable VLM Self-Evolution via Multi-Agent Self-Play

HiPRAG: Hierarchical Process Rewards for Efficient Agentic Retrieval Augmented Generation

GEAR: A $G$eneral $E$valuation Framework for $A$bductive $R$easoning

Exploring the Entropy Mechanism in LLM Agents On-policy Optimization

Understanding and Mitigating Numerical Sources of Nondeterminism in LLM Inference

Efficient Self-Improvement in Multimodal Large Language Models: A Model-Level Judge-Free Approach