Haochen Zhang

Q-Learning with Fine-Grained Gap-Dependent Regret

Gap-Dependent Bounds for Q-Learning using Reference-Advantage Decomposition

Regret-Optimal Q-Learning with Low Cost for Single-Agent and Federated Reinforcement Learning

Gap-Dependent Bounds for Federated $Q$-Learning

Federated $Q$-Learning with Reference-Advantage Decomposition: Almost Optimal Regret and Logarithmic Communication Cost