Puheng Li (李朴恒)

Puheng Li (李朴恒)

Ph.D. Student

Stanford

I'm a third-year PhD student in the Department of Statistics at Stanford University, where I'm fortunate to be advised by Prof. Emmanuel Candès and Prof. Renyuan Xu. My research interest lies broadly in leveraging statistical methods to guide modern machine learning and make them more interpretable, reliable, efficient and accurate with statistical guarantees.

Previously, I received my bachelor's degree in statistics from School of Mathematical Sciences at Peking University. I had the pleasure of working with Prof. Lei Wu, Prof. Weijie Su, Prof. Linjun Zhang, and Prof. James Zou during my undergraduate study. Please feel free to contact me if you are interested in my research!

Contact
puhengli AT stanford DOT edu
Education

  • Ph.D. in Statistics, 2023-

    Department of Statistics, Stanford University

  • B.S. in Statistics, 2023

    School of Mathematical Sciences, Peking University

  • High School Diploma (Mathematics), 2019

    Shanghai High School

(*: Equal contribution.)

Publications

Jiaqi Han, Juntong Shi, Puheng Li, Haotian Ye, Qiushan Guo, Stefano Ermon. Adaptive Spectral Feature Forecasting for Diffusion Sampling Acceleration. The IEEE/CVF Conference on Computer Vision and Pattern Recognition 2026 (CVPR 2026).

arXiv Project Page

Haotian Ye, Qiyuan He, Jiaqi Han, Puheng Li, Jiaojiao Fan, Zekun Hao, Fitsum Reda, Yogesh Balaji, Huayu Chen, Sheng Liu, Angela Yao, James Zou, Stefano Ermon, Haoxiang Wang, Ming-Yu Liu. InfoTok: Adaptive Discrete Video Tokenizer via Information-Theoretic Compression. The Fourteenth International Conference on Learning Representations (ICLR 2026, Oral).

arXiv Project Page

Puheng Li, Tijana Zrnic, Emmanuel Candès. Robust Sampling for Active Statistical Inference. The Thirty-Ninth Annual Conference on Neural Information Processing Systems (NeurIPS 2025).

arXiv Poster

Jiaqi Han, Haotian Ye, Puheng Li, Minkai Xu, James Zou, Stefano Ermon. CHORDS: Diffusion Sampling Accelerator with Multi-core Hierarchical ODE Solvers. International Conference on Computer Vision (ICCV 2025).

arXiv Project Page

Keyao Zhan*, Puheng Li*, Lei Wu. Analyzing the Role of Permutation Invariance in Linear Mode Connectivity. The 28th International Conference on Artificial Intelligence and Statistics (AISTATS 2025).

arXiv Poster

Puheng Li*, Zhong Li*, Huishuai Zhang, Jiang Bian. On the Generalization Properties of Diffusion Models. The Thirty-Seventh Annual Conference on Neural Information Processing Systems (NeurIPS 2023).

arXiv Poster

Puheng Li, James Zou, Linjun Zhang. FaiREE: Fair Classification with Finite-Sample and Distribution-Free Guarantee. The Eleventh International Conference on Learning Representations (ICLR 2023).

arXiv Poster

Preprints

Wenbo Chen*, Puheng Li*, Mengyang Liu*, Weijie Su, Tianpei Xie. MARS: Margin-Adversarial Risk-controlled Stopping for Parallel LLM Test-time Scaling.

arXiv

Jiaqi Han*, Puheng Li*, Qiushan Guo, Renyuan Xu, Stefano Ermon, Emmanuel J. Candès. One-Step Generative Modeling via Wasserstein Gradient Flows.

arXiv Project Page

Haotian Ye, Kaiwen Zheng, Jiashu Xu, Puheng Li, Huayu Chen, Jiaqi Han, Sheng Liu, Qinsheng Zhang, Hanzi Mao, Zekun Hao, Prithvijit Chattopadhyay, Dinghao Yang, Liang Feng, Maosheng Liao, Junjie Bai, Ming-Yu Liu, James Zou, Stefano Ermon. Data-regularized Reinforcement Learning for Diffusion Models at Scale.

arXiv Project Page

Zhanran Lin*, Puheng Li*, Lei Wu. Exploring Neural Network Landscapes: Star-Shaped and Geodesic Connectivity.

arXiv

Talks

"One-Step Generative Modeling via Wasserstein Gradient Flows."
Aug. 6, 2026 · Joint Statistical Meetings, Boston.

Link

"On the Reliability and Efficiency of Generative AI."
Aug. 22, 2025 · Cubist Systematic Strategies.
"On the Generalization Properties of Diffusion Models."
Jan. 1, 2024 · MIT Learning on Graphs & Geometry (LoGG) Reading Group.

Link

Work Experience

Quantitative Analyst Intern, The D. E. Shaw Group (Jun. 2026 - Sep. 2026)
Applied Scientist Intern, Amazon (Jun. 2025 - Sep. 2025)
  • Worked on LLMs for fraud detection and prevention.
Research Scientist Intern, Machine Learning Group, Microsoft Research Asia (Mar. 2023 - Jun. 2023)

Teaching & Service

Teaching Assistant, Stanford University
  • STATS 203 Regression Models and Analysis of Variance 2025-2026 Spring
  • STATS 305B Applied Statistics II 2025-2026 Winter
  • STATS 117 Introduction to Probability Theory 2025-2026 Autumn
  • STATS 208 Resampling Methods: Bootstrap, Cross Validation and Beyond 2024-2025 Spring
  • STATS 207/307 Time Series Analysis 2024-2025 Autumn
  • STATS 191 Introduction to Applied Statistics 2023-2024 Summer
  • STATS 207/307 Introduction to Time Series Analysis 2023-2024 Spring
  • STATS 100 Mathematics of Sports 2023-2024 Winter
Reviewer

ICLR 2025, NeurIPS 2025 (Main Conference + GenAI in Finance Workshop) & 2026, CVPR 2026, ISIT 2026

Honors & Awards

  • 2025 Finalist, Cubist PhD Fellowship
  • 2023 Stars of Tomorrow, Microsoft Research Asia
  • 2023 Excellent Graduate (PKU & Beijing), equiv. to Summa Cum Laude
  • 2022 Leo Koguan Scholarship
  • 2022, 2021, 2020 Merit Student of Peking University
  • 2021 Peking University Talent Pool Scholarship
  • 2020 Yizheng Alumni Scholarship
  • 2018 Gold Medal at the 34th Chinese Mathematical Olympiad (CMO)
  • 2017 Silver Medal at the 33rd Chinese Mathematical Olympiad (CMO)

Miscellaneous

I love singing, especially solo and a cappella. I was honored one of the Top 10 Singers at Peking University.

Find some of my performances on my Bilibili!

I enjoy playing the guitar, chess, Go, table tennis, and pool. I also had fun providing voice-over for radio dramas and school promotional videos.

Try out a Gomoku Game with the robot I programmed in college!
(Bot ID: 6363ccda832a161fc44b43d3)