Publications

Publications by categories in reversed chronological order. The authors of most papers are ordered alphabetically.

2024

  1. Neurips 24
    Private Stochastic Convex Optimization with Heavy Tails: Near-Optimality from Simple Reductions
    Hilal Asi, Daogao Liu, and Kevin Tian
    arXiv preprint arXiv:2406.02789, 2024
  2. Neurips 24
    Private Online Learning via Lazy Algorithms
    Hilal Asi, Tomer Koren, Daogao Liu, and Kunal Talwar
    arXiv preprint arXiv:2406.03620, 2024
  3. Neurips 24
    Faster Algorithms for User-Level Private Stochastic Convex Optimization
    Andrew Lowy, Daogao Liu, and Hilal Asi
    arXiv preprint arXiv:2410.18391, 2024
  4. COLM 24
    Mind the Privacy Unit! User-Level Differential Privacy for Language Model Fine-Tuning
    Lynn Chua, Badih Ghazi, Yangsibo Huang, Pritish Kamath, Daogao Liu, Pasin Manurangsi, Amer Sinha, and Chiyuan Zhang
    arXiv preprint arXiv:2406.14322, 2024
  5. ICML 24
    Private gradient descent for linear regression: Tighter error bounds and instance-specific uncertainty estimation
    Gavin Brown, Krishnamurthy Dvijotham, Georgina Evans, Daogao Liu, Adam Smith, and Abhradeep Thakurta
    arXiv preprint arXiv:2402.13531, 2024
  6. AISTATS 24
    User-level differentially private stochastic convex optimization: Efficient algorithms with optimal rates
    Daogao Liu, and Hilal Asi
    In International Conference on Artificial Intelligence and Statistics, 2024

2023

  1. ICLR 24
    Detecting Pretraining Data from Large Language Models
    Weijia Shi, Anirudh Ajith, Menthou Xia, Yangsibo Huang, Daogao Liu, Terra Blevin, Danqi Chen, and Luke Zettlemoyer
    arXiv preprint arXiv:2310.16789, 2023
    arXiv
  2. Neurips 23 Spotlight
    Private (Stochastic) Non-Convex Optimization Revisited: Second-Order Stationary Points and Excess Risks
    Arun Ganesh, Daogao Liu, Sewoong Oh, and Abhradeep Thakurta
    arXiv preprint arXiv:2302.09699, 2023
    arXiv
  3. FOCS 23
    ReSQueing Parallel and Private Stochastic Convex Optimization
    Yair Carmon, Arun Jambulapati, Yujia Jin, Yin Tat Lee, Daogao Liu, Aaron Sidford, and Kevin Tian
    In 64th IEEE Symposium on Foundations of Computer Science (FOCS), 2023
    arXiv
  4. COLT 23
    Algorithmic Aspects of the Log-Laplace Transform and a Non-Euclidean Proximal Sampler
    Sivakanth Gopi, Yin Tat Lee, Daogao Liu, Ruoqi Shen, and Kevin Tian
    In Conference on Learning Theory, 2023
    arXiv
  5. STOC 23
    Pandora Box Problem with Nonobligatory Inspection: Hardness and Improved Approximation Algorithms
    Hu Fu, Jiawei Li, and Daogao Liu
    In Proceedings of the 55th Annual ACM Symposium on Theory of Computing (STOC), 2023
    arXiv
  6. SODA 23
    Private Convex Optimization in General Norms
    Sivakanth Gopi, Yin Tat Lee, Daogao Liu, Ruoqi Shen, and Kevin Tian
    In Proceedings of the 2023 Annual ACM-SIAM Symposium on Discrete Algorithms (SODA), 2023
    arXiv
  7. SODA 23
    Super-resolution and Robust Sparse Continuous Fourier Transform in Any Constant Dimension: Nearly Linear Time and Sample Complexity
    Yaonan Jin, Daogao Liu, and Zhao Song
    In Proceedings of the 2023 Annual ACM-SIAM Symposium on Discrete Algorithms (SODA), 2023
    arXiv
  8. ICLR 23
    Augmentation with Projection: Towards an Effective and Efficient Data Augmentation Paradigm for Distillation
    Ziqi Wang, Yuexin Wu, Frederick Liu, Daogao Liu, Le Hou, Hongkun Yu, Jing Li, and Heng Ji
    arXiv preprint arXiv:2210.11768, 2023
    arXiv

2022

  1. Neurips 22
    When Does Differentially Private Learning Not Suffer in High Dimensions?
    Xuechen Li*, Daogao Liu*, Tatsunori Hashimoto, Huseyin A Inan, Janardhan Kulkarni, Yin Tat Lee, and Abhradeep Guha Thakurta
    arXiv preprint arXiv:2207.00160, 2022
    arXiv
  2. COLT 22
    Private convex optimization via exponential mechanism
    Sivakanth Gopi, Yin Tat Lee, and Daogao Liu
    In Conference on Learning Theory, 2022
    arXiv
  3. COLT 22
    Better private algorithms for correlation clustering
    Daogao Liu
    In Conference on Learning Theory, 2022
    arXiv
  4. SODA 22
    Multi-token Markov Game with Switching Costs∗
    Jian Li, and Daogao Liu
    In Proceedings of the 2022 Annual ACM-SIAM Symposium on Discrete Algorithms (SODA), 2022
    arXiv

2021

  1. Neurips 21 Spotlight
    Private non-smooth erm and sco in subquadratic steps
    Janardhan Kulkarni, Yin Tat Lee, and Daogao Liu
    Advances in Neural Information Processing Systems, 2021
    arXiv

2020

  1. ITCS 20
    Algorithms and Adaptivity Gaps for Stochastic k-TSP
    Haotian Jiang, Jian Li, Daogao Liu, and Sahil Singla
    In 11th Innovations in Theoretical Computer Science Conference (ITCS 2020), 2020
    arXiv

2019

  1. COCOON 19
    More Efficient Algorithms for Stochastic Diameter and Some Unapproximated Problems in Metric Space
    Daogao Liu
    In Computing and Combinatorics: 25th International Conference, COCOON 2019, Xi’an, China, July 29–31, 2019, Proceedings, 2019
    PDF

Manuscript

2024

  1. Adaptive Batch Size for Privately Finding Second-Order Stationary Points
    Daogao Liu, and Kunal Talwar
    arXiv preprint arXiv:2410.07502, 2024
  2. Improved Sample Complexity for Private Nonsmooth Nonconvex Optimization
    Guy Kornowski, Daogao Liu, and Kunal Talwar
    arXiv preprint arXiv:2410.05880, 2024
  3. Unlearn and Burn: Adversarial Machine Unlearning Requests Destroy Model Accuracy
    Yangsibo Huang, Daogao Liu, Lynn Chua, Badih Ghazi, Pritish Kamath, Ravi Kumar, Pasin Manurangsi, Milad Nasr, Amer Sinha, and Chiyuan Zhang
    arXiv preprint arXiv:2410.09591, 2024
  4. Muse: Machine unlearning six-way evaluation for language models
    Weijia Shi, Jaechan Lee, Yangsibo Huang, Sadhika Malladi, Jieyu Zhao, Ari Holtzman, Daogao Liu, Luke Zettlemoyer, Noah A Smith, and Chiyuan Zhang
    arXiv preprint arXiv:2407.06460, 2024