Baoxiang Wang

3 papers The Chinese University of Hong Kong

• ADG: Ambient Diffusion-Guided Dataset Recovery for Corruption-Robust Offline Reinforcement Learning
• Scalable Exploration via Ensemble++
• Uncoupled and Convergent Learning in Monotone Games under Bandit Feedback