avatar

I am a PhD student majoring in computer science at the School of Computing, National University of Singapore, advised by Reza Shokri. Here is my CV (until Jan 17, 2026).

I work on privacy and data protection in machine learning. My research analyzes how individual training examples affects learned model behaviors, with the goal of making learning more controllable and privacy-perserving. My long-term research goal is to enhance learning performance on real-world datasets that are long-tailed, noisy, and heterogeneous, especially when they contain privacy-sensitive information.

My current research focus on (1) developing improved algorithms for learning with differential privacy guarantees, (2) enabling controllable learning of distinct capabilities through strategic data selection and parameter separation. Currently, I’m a research intern at Apple ML Research working to boost fact memorization and memory-intensive reasoning for language models trained on long-tailed, heavily mixed data.

I’m a recipient of the 2024 Apple Scholars in AI/ML PhD fellowship and the 2023-2024 Google PhD Fellowship in security and privacy. Previously, I had a wonderful time interning at Apple ML Research in 2024 Spring and Azure Research - Microsoft Research in 2023 summer. I received my B.S. degree in computational mathematics at University of Science and Technology of China, where I had a memorable time.

Selected Publications

(* denotes equal contribution)

(see Google Scholar for complete publications)

  1. Cram Less to Fit More: Training Data Pruning Improves Memorization of Facts [Paper]
    Jiayuan Ye, Vitaly Feldman, Kunal Talwar

  2. Instance-Optimality for Private KL Distribution Estimation [Paper]
    Jiayuan Ye, Vitaly Feldman, Kunal Talwar
    In Advances in Neural Information Processing Systems (NeurIPS) 2025
    (Accepted as spotlight, among ~3% of submissions)
    Also Presented at the Theory and Practice of Differential Privacy (TPDP) 2025

  3. How much of my dataset did you use? Quantitative Data Usage Inference in Machine Learning [Paper]
    Yao Tong*, Jiayuan Ye*, Sajjad Zarifzadeh, Reza Shokri
    In International Conference on Learning Representations (ICLR) 2025
    (Acccepted as oral, among ~2% of submissions)

  4. Leave-one-out Distinguishability in Machine Learning [Paper] [Code]
    Jiayuan Ye, Anastasia Borovykh, Soufiane Hayou, Reza Shokri
    In International Conference on Learning Representations (ICLR) 2024
    Also Presented at the Symposium on Foundations of Responsible Computing (FORC) 2024

  5. Initialization Matters: Privacy-Utility Analysis of Overparameterized Neural Networks [Paper] [Poster]
    Jiayuan Ye, Zhenyu Zhu, Fanghui Liu, Reza Shokri, Volkan Cevher
    In Advances in Neural Information Processing Systems (NeurIPS) 2023
    Also Presented at the Theory and Practice of Differential Privacy (TPDP) 2023

  6. Differentially Private Learning Needs Hidden State (Or Much Faster Convergence) [Paper] [Poster] [Talk]
    Jiayuan Ye, Reza Shokri
    In Advances in Neural Information Processing Systems (NeurIPS) 2022
    Also Presented at the Symposium on Foundations of Responsible Computing (FORC) 2022

  7. Enhanced Membership Inference Attacks against Machine Learning Models [Paper] [Slides] [Code]
    Jiayuan Ye, Aadyaa Maddi, Sasi Kumar Murakonda, Vincent Bindschaedler, Reza Shokri
    In the ACM Conference on Computer and Communications Security (CCS) 2022
    (Among top 10 most cited papers published in security conferences in 2022. Link: \url{https://mlsec.org/topnotch/sec_2020s.html})

  8. Differential Privacy Dynamics of Langevin Diffusion and Noisy Gradient Descent [Paper] [Talk] [Slides] [Poster]
    Rishav Chourasia*, Jiayuan Ye*, Reza Shokri
    In Advances in Neural Information Processing Systems (NeurIPS) 2021
    (Accepted as spotlight, among ~3% of submissions)

Professional Experiences

  • Conference & Workshop Program Commitee/Reviewer: NeurIPS 2022, 2023, 2024, 2025; ICLR 2023, 2024, 2025, 2026; ICML 2023, 2024, 2025; AISTATS 2023, 2025; ACM CCS 2024; IEEE SaTML 2025, 2026; TPDP 2025, 2026; PPAI-2022; FL-ICML 2023; PRIVATE ML @ ICLR 2024; SYNTHDATA @ ICLR 2025; DATA-FM @ ICLR 2025.
  • Journal Reviewer: JMLR (2022), SICOMP (2023)
  • Conference & Workshop sub-reviewer: IEEE S&P 2020, 2021, 2022, 2023, 2024; PPAI 2021; ACM CCS 2021, 2022, 2023