Research

* denotes equal contribution and ** denotes alphabet order. An up-to-date list is available on Google Scholar.

Large Language Models (LLMs)

  1. A statistical framework of watermarks for large language models: Pivot, detection efficiency and optimal rules
    Xiang Li, Feng Ruan, Huiyuan Wang, Qi Long, and Weijie Su
    The Annals of Statistics, 2025
    arXiv HTML Code Slides
  2. Robust detection of watermarks in large language models under human edits
    Xiang Li, Feng Ruan, Huiyuan Wang, Qi Long, and Weijie Su
    arXiv preprint arXiv:2411.13868, 2024
    arXiv Code Slides
  3. Debiasing watermarks for large language models via maximal coupling
    Yangxinyu Xie, Xiang Li, Tanwi Mallick, Weijie Su, and Ruixun Zhang
    arXiv preprint arXiv:2411.11203, 2024
    arXiv Code

Statistical Inference

  1. Online statistical inference for nonlinear stochastic approximation with Markovian data
    Xiang Li, Jiadong Liang, and Zhihua Zhang
    arXiv preprint arXiv:2302.07690, 2023
    arXiv
  2. A statistical analysis of Polyak-Ruppert averaged Q-learning
    Xiang Li, Wenhao Yang, Jiadong Liang, Zhihua Zhang, and Michael I Jordan
    In International Conference on Artificial Intelligence and Statistics, 2023
    arXiv Code Poster Slides Video
  3. Statistical estimation and online inference via Local SGD
    Xiang Li, Jiadong Liang, Xiangyu Chang, and Zhihua Zhang
    In Conference on Learning Theory, 2022
    HTML Slides
  4. Convergence and inference of Stream SGD, with applications to queueing systems and inventory control
    Xiang Li, Jiadong Liang, Xinyun Chen, and Zhihua Zhang
    arXiv preprint arXiv:2309.09545, 2023
    arXiv Code
  5. Uncertainty quantification of data shapley via statistical inference
    Mengmeng Wu, Zhihong Liu, Xiang Li, Ruoxi Jia, and Xiangyu Chang
    arXiv preprint arXiv:2407.19373, 2024
    arXiv
  6. Statistical analysis of Karcher means for random restricted PSD matrices
    Hengchao Chen, Xiang Li, and Qiang Sun
    In International Conference on Artificial Intelligence and Statistics, 2023
    arXiv

Stochastic Approximation

  1. Finite-time decoupled convergence in nonlinear two-time-scale stochastic approximation
    Yuze Han, Xiang Li, and Zhihua Zhang
    arXiv preprint arXiv:2401.03893, 2024
    arXiv
  2. Decoupled functional central limit theorems for two-time-scale stochastic approximation
    Yuze Han, Xiang Li, Jiadong Liang, and Zhihua Zhang
    arXiv preprint arXiv:2412.17070, 2024
    arXiv
  3. Asymptotic behaviors and phase transitions in projected stochastic approximation: A jump diffusion approach
    Jiadong Liang, Yuze Han, Xiang Li, and Zhihua Zhang
    arXiv preprint arXiv:2304.12953, 2023
    Extended version of the conference paper: Asymptotic behaviors of projected stochastic approximation: A jump diffusion perspective
    arXiv
  4. Asymptotic behaviors of projected stochastic approximation: A jump diffusion perspective
    Jiadong Liang, Yuze Han, Xiang Li, and Zhihua Zhang
    In Advances in Neural Information Processing Systems, Spotlight, 2022
    HTML Slides
  5. Do subsampled Newton methods work for high-dimensional data?
    Xiang Li, Shusen Wang, and Zhihua Zhang
    In AAAI Conference on Artificial Intelligence, 2020
    HTML

Federated Learning

  1. On the convergence of FedAvg on non-iid data
    Xiang Li*, Kaixuan Huang*, Wenhao Yang*, Shusen Wang, and Zhihua Zhang
    In International Conference on Learning Representations, Oral Presentation, 2020
    HTML Code Poster Slides
  2. A random projection approach to personalized federated learning: Enhancing communication efficiency, robustness, and fairness
    Yuze Han**, Shiyun Lin**, Xiang Li**, and Zhihua Zhang**
    Journal of Machine Learning Research, 2024
    Extended version of the conference paper: Personalized federated learning towards communication efficiency, robustness and fairness
    HTML
  3. Fedpower: Privacy‑preserving distributed eigenspace estimation
    Xiao Guo, Xiang Li, Xiangyu Chang, Shusen Wang, and Zhihua Zhang
    Machine Learning, 2024
    Extended version of the conference paper: Communication-efficient distributed SVD via local power iterations
    arXiv HTML
  4. Personalized federated learning towards communication efficiency, robustness and fairness
    Shiyun Lin*, Yuze Han*, Xiang Li, and Zhihua Zhang
    In Neural Information Processing Systems, 2022
    HTML Code Poster Slides
  5. Communication-efficient distributed SVD via local power iterations
    Xiang Li, Shusen Wang, Kun Chen, and Zhihua Zhang
    In International Conference on Machine Learning, 2021
    HTML Code Poster Slides
  6. Communication efficient decentralized training with multiple local updates
    Xiang Li, Wenhao Yang, Shusen Wang, and Zhihua Zhang
    arXiv preprint arXiv:1910.09126, 2019
    arXiv

Online Decision Making

  1. Variance-aware decision making with linear function approximation with heavy-tailed rewards
    Xiang Li, and Qiang Sun
    Transactions on Machine Learning Research, 2024
    arXiv HTML
  2. A regularized approach to sparse optimal policy in reinforcement learning
    Wenhao Yang*, Xiang Li*, and Zhihua Zhang
    In Advances in Neural Information Processing Systems, 2019
    HTML
  3. Finding near optimal policies via reducive regularization in Markov decision processes
    Wenhao Yang, Xiang Li, Guangzeng Xie, and Zhihua Zhang
    In Workshop on Reinforcement Learning Theory, ICML, 2021
    HTML