← Search

Ziqing Mai

2 accepted papers

2024

Bayesian Design Principles for Offline-to-Online Reinforcement Learning

ICML 2024poster

Offline reinforcement learning (RL) is crucial for real-world applications where exploration can be costly or unsafe. However, offline learned policies are often suboptimal, and further online fine-tuning is required. In this paper, we tackle the fundamental dilemma of offline-to-online fine-tuning:…

2023

Unsupervised Behavior Extraction via Random Intent Priors

NeurIPS 2023poster

Reward-free data is abundant and contains rich prior knowledge of human behaviors, but it is not well exploited by offline reinforcement learning (RL) algorithms. In this paper, we propose UBER, an unsupervised approach to extract useful behaviors from offline reward-free datasets via diversified re…

Cited by 8SourcePDFScholar