Weiyan Shi

PolySkill: Learning Generalizable Skills Through Polymorphic Abstraction For Continual Learning

SPIRAL: Self-Play on Zero-Sum Games Incentivizes Reasoning via Multi-Agent Multi-Turn Reinforcement Learning

GEM: A Gym for Generalist LLMs

Can Aha Moments Be Fake? Identifying True and Decorative Thinking Steps in Chain-of-Thought

Verbalized Sampling: How to Mitigate Mode Collapse and Unlock LLM Diversity

LLMs Encode Harmfulness and Refusal Separately

Distilling an End-to-End Voice Assistant Without Instruction Training Data