Zongzhang Zhang

Learning Disentangled Multi-Agent World Model for Decentralized Control

EvaLearn: Quantifying the Learning Capability and Efficiency of LLMs via Sequential Problem Solving

Focus-Then-Reuse: Fast Adaptation in Visual Perturbation Environments

Behavior-Regularized Diffusion Policy Optimization for Offline Reinforcement Learning

Q-Adapter: Customizing Pre-trained LLMs to New Preferences with Forgetting Mitigation

Reinforced In-Context Black-Box Optimization

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

Hindsight Preference Learning for Offline Preference-based Reinforcement Learning

Diffusion-Guided Safe Policy Optimization From Cost-Label-Free Offline Dataset