Lei Han

Tackling Data Corruption in Offline Reinforcement Learning via Sequence Modeling

Learning 3D Persistent Embodied World Models

STAR: Efficient Preference-based Reinforcement Learning via Dual Regularization

VidEgoThink: Assessing Egocentric Video Understanding Capabilities for Embodied AI

Cracking the Collective Mind: Adversarial Manipulation in Multi-Agent Systems

Task Adaptation from Skills: Information Geometry, Disentanglement, and New Objectives for Unsupervised Reinforcement Learning

Towards Robust Offline Reinforcement Learning under Diverse Data Corruption

HumanVLA: Towards Vision-Language Directed Object Rearrangement by Physical Humanoid

Self-playing Adversarial Language Game Enhances LLM Reasoning

Toward Self-Improvement of LLMs via Imagination, Searching, and Criticizing

Efficient and scalable reinforcement learning via hypermodel

SEER: Towards Efficient Preference-based Reinforcement Learning via Aligned Experience Estimation