publications

2026

  1. Temporal Gradient Inversion for Private Trajectory Reconstruction in Embodied Reinforcement Learning
    Sudip Bhujel, Shanghao Shi, Ruiquan Huang, Ning Zhang, and Yang Xiao
    In Advances in Neural Information Processing Systems (NeurIPS), 2026
    Accepted
    arXiv
  2. Flow Matching for Offline Reinforcement Learning with Discrete Actions
    Fairoz Nower Khan, Nabuat Zaman Nahim, Ruiquan Huang, Haibo Yang, and Peizhong Ju
    In Advances in Neural Information Processing Systems (NeurIPS), 2026
    Accepted
    arXiv
  3. Breaking the Computational Barrier: Provably Efficient Actor-Critic for Low-Rank MDPs
    Ruiquan Huang, Donghao Li, Yingbin Liang, and Jing Yang
    International Conference on Machine Learning (ICML), 2026
    arXiv

2025

  1. Robust Offline Reinforcement Learning for Non-Markovian Decision Processes
    Ruiquan Huang, Yingbin Liang, and Jing Yang
    IEEE Transactions on Information Theory, 2025
    DOI arXiv
  2. Augmenting Online RL with Offline Data is All You Need: A Unified Hybrid RL Algorithm Design and Analysis
    Ruiquan Huang, Donghao Li, Chengshuai Shi, Cong Shen, and Jing Yang
    In Conference on Uncertainty in Artificial Intelligence (UAI), 2025
  3. How Transformers Learn Regular Language Recognition: A Theoretical Study on Training Dynamics and Implicit Bias
    Ruiquan Huang, Yingbin Liang, and Jing Yang
    In International Conference on Machine Learning (ICML), 2025

2024

  1. Federated Online Prediction from Experts with Differential Privacy: Separations and Regret Speed-ups
    Fengyu Gao, Ruiquan Huang, and Jing Yang
    In The Thirty-eighth Annual Conference on Neural Information Processing Systems (NeurIPS), 2024
  2. Temporal-distributed backdoor attack against video based action recognition
    Xi Li, Songhe Wang, Ruiquan Huang, Mahanth Gowda, and George Kesidis
    In Proceedings of the AAAI Conference on Artificial Intelligence, 2024
  3. Provably efficient ucb-type algorithms for learning predictive state representations
    Ruiquan Huang, Yingbin Liang, and Jing Yang
    12th International Conference on Learning Representations (ICLR), 2024
  4. Non-asymptotic Convergence of Training Transformers for Next-token Prediction
    Ruiquan Huang, Yingbin Liang, and Jing Yang
    The Thirty-eighth Annual Conference on Neural Information Processing Systems (NeurIPS), 2024
  5. Provable Benefits of Multi-task RL under Non-Markovian Decision Making Processes
    Ruiquan Huang, Yuan Cheng, Jing Yang, Vincent Tan, and Yingbin Liang
    12th International Conference on Learning Representations (ICLR), 2024
  6. Towards General Function Approximation in Nonstationary Reinforcement Learning
    Songtao Feng, Ming Yin, Ruiquan Huang, Yu-Xiang Wang, Jing Yang, and Yingbin Liang
    IEEE Journal on Selected Areas in Information Theory, 2024

2023

  1. Federated linear contextual bandits with user-level differential privacy
    Ruiquan Huang, Huanyu Zhang, Luca Melis, Milan Shen, Meisam Hejazinia, and Jing Yang
    In International Conference on Machine Learning (ICML), 2023
  2. Improved sample complexity for reward-free reinforcement learning under low-rank mdps
    Yuan Cheng, Ruiquan Huang, Jing Yang, and Yingbin Liang
    11th International Conference on Learning Representations (ICLR), 2023
  3. Non-stationary reinforcement learning under general function approximation
    Songtao Feng, Ming Yin, Ruiquan Huang, Yu-Xiang Wang, Jing Yang, and Yingbin Liang
    In International Conference on Machine Learning (ICML), 2023
  4. Safe exploration incurs nearly no additional sample complexity for reward-free rl
    Ruiquan Huang, Jing Yang, and Yingbin Liang
    11th International Conference on Learning Representations (ICLR), 2023
  5. FLORAS: Differentially private wireless federated learning using orthogonal sequences
    Xizixiang Wei, Tianhao Wang, Ruiquan Huang, Cong Shen, Jing Yang, and H Vincent Poor
    In ICC 2023-IEEE International Conference on Communications, 2023
  6. Near-optimal conservative exploration in reinforcement learning under episode-wise constraints
    Donghao Li, Ruiquan Huang, Cong Shen, and Jing Yang
    In International Conference on Machine Learning, 2023

2022

  1. Cascading bandits with two-level feedback
    Duo Cheng, Ruiquan Huang, Cong Shen, and Jing Yang
    In 2022 IEEE International Symposium on Information Theory (ISIT), 2022

2021

  1. Federated linear contextual bandits
    Ruiquan Huang, Weiqiang Wu, Jing Yang, and Cong Shen
    Advances in neural information processing systems, 2021