Hongming Zhang

Principled Fast and Meta Knowledge Learners for Continual Reinforcement Learning

Generalizable Opponent Exploitation in LLM Agents via Mixed Best-Responses Training

FPDou: Mastering DouDizhu with Fictitious Play

Principled Fast and Meta Knowledge Learners for Continual Reinforcement Learning

STAR: Efficient Preference-based Reinforcement Learning via Dual Regularization

Double Check My Desired Return: Transformer with Value Validation for Offline RL