Pengyuan Wang

Off-Policy Token Clipped Supervised Fine-Tuning Yields a Robust Cold-Start

Long CoT In-Context Learning Can Empower Pre-trained LLMs

xVerify: Efficient Answer Verifier for Reasoning Model Evaluations

Semantic Temporal Abstraction via Vision-Language Model Guidance for Efficient Reinforcement Learning

Controlling Large Language Model with Latent Action

Multi-Agent Imitation by Learning and Sampling from Factorized Soft Q-Function

Whale-X: Learning Scalable Embodied World Models with Enhanced Generalizability