Chongxuan Li（李崇轩）

Assistant Professor
Gaoling School of AI
Renmin University of China
Email: chongxuanli [at] ruc (dot) edu (dot) cn

[Google Scholar] [Github] [Profiles-CN]

Research Topics
  • Deep Generative Model
  • Diffusion Probabilistic Model
  • Learning with Limited Data
  • Theory of Generative Model
  • Generative AI for Science

I am a tenure-tack assistant professor at Gaoling School of AI, Renmin University of China. I am leading the machine learning group at RUC [Github]. The long term goal of my research is building a machine that can work well on limited, complex and uncertain data emerged in real world applications. To this end, my research foucses on statistically and computationally efficient probabilistic models, algorithms and theory.

Before joining RUC, I was a Post Doc researcher in TSAIL at Tsinghua University, worked with Prof. Jun Zhu. I recieved my Ph.D. in the Department of Computer Science and Technology at Tsinghua University, supervised by Prof. Bo Zhang and Jun Zhu. During my Ph.D., I had spent a great year at AMLab, working with Prof. Max Welling. I received the B.E. degree from Institute for Interdiscriplinary Information Sciences at Tsinghua University in 2014.

Positions available for self-motivated Ph.D and master students. :-)

ML Group @ RUC

PhD Students
  • Fan Bao (with Jun Zhu)
  • Min Zhao (with Jun Zhu)
  • Yong Zhong
  • Shen Nie
  • Zebin You
  • Yidong Ouyang (visiting, with Guang Cheng and Liyan Xie)
Master Students
  • Hanzhong Guo (visiting)
Undergraduate
  • Rongzhen Wang
  • Kaiwen Xue
  • Chenyu Zheng

Teaching

  • Probability and Randomized Algorithms (Graduate, Autumn, since 2021)
  • Deep Generative Models: Principles and Applications (Graduate, Spring, since 2022)
  • Probabilistic Graphical Models: Principles and Applications (Undergraduate, Spring, since 2023)

Publications (* indicates equal contribution and † indicates correspondence.)

    Preprints
  • One Transformer Fits All Distributions in Multi-Modal Diffusion at Scale
    Fan Bao, Shen Nie, Kaiwen Xue, Chongxuan Li, Shi Pu, Yaole Wang, Gang Yue, Yue Cao, Hang Su, Jun Zhu
    [Paper] [Code] [Hugging Face] [Colab]
  • Diffusion Models and Semi-Supervised Learners Benefit Mutually with Few Labels
    Zebin You, Yong Zhong, Fan Bao, Jiacheng Sun, Chongxuan Li†, Jun Zhu
    [Paper] [Code] [Demo]
  • Revisiting Discriminative vs. Generative Classifiers: Theory and Implications
    Chenyu Zheng, Guoqiang Wu, Fan Bao, Yue Cao, Chongxuan Li†, Jun Zhu
    [Paper] [Code]
    • 2023
  • All are Worth Words: a ViT Backbone for Diffusion Models
    Fan Bao, Shen Nie, Kaiwen Xue, Yue Cao, Hang Su, Chongxuan Li†, Jun Zhu†
    The IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR), 2023
  • Equivariant Energy-Guided SDE for Inverse Molecular Design
    Fan Bao, Min Zhao, Zhongkai Hao, Peiyao Li, Chongxuan Li†, Jun Zhu†
    International Conference on Learning Representations (ICLR) 2023
  • Deep Generative Modeling on Limited Data with Regularization by Nontransferable Pre-trained Models
    Yong Zhong, Hongtao Liu, Xiaodong Liu, Fan Bao, Weiran Shen†, Chongxuan Li†
    International Conference on Learning Representations (ICLR) 2023
    • 2022
  • Why Are Conditional Generative Models Better Than Unconditional Ones?
    Fan Bao, Chongxuan Li†, Jiacheng Sun, Jun Zhu†
    Score-based Model workshop @ Advances in Neural Information Processing Systems (SBM@NeurIPS) 2022
  • EGSDE: Unpaired Image-to-Image Translation via Energy-Guided Stochastic Differential Equations
    Min Zhao, Fan Bao, Chongxuan Li†, Jun Zhu†
    Advances in Neural Information Processing Systems (NeurIPS) 2022
  • DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps
    Cheng Lu, Yuhao Zhou, Fan Bao, Jianfei Chen, Chongxuan Li, Jun Zhu
    Advances in Neural Information Processing Systems (NeurIPS) 2022
    Oral Presentation
  • Fast Lossless Neural Compression with Integer-Only Discrete Flows
    Siyu Wang, Jianfei Chen, Chongxuan Li, Jun Zhu, Bo Zhang
    International Conference on Machine Learning (ICML), 2022
  • Maximum Likelihood Training for Score-based Diffusion ODEs by High Order Denoising Score Matching
    Cheng Lu, Kaiwen Zheng, Fan Bao, Chongxuan Li, Jianfei Chen, Jun Zhu
    International Conference on Machine Learning (ICML), 2022
  • Estimating the Optimal Covariance with Imperfect Mean in Diffusion Probabilistic Models
    Fan Bao, Chongxuan Li†, Jiacheng Sun, Jun Zhu†, Bo Zhang
    International Conference on Machine Learning (ICML), 2022
    [Code]
  • Memory Replay with Data Compression for Continual Learning
    Liyuan Wang, Xingxing Zhang, Kuo Yang, Longhui Yu, Chongxuan Li†, Lanqing Hong, Shifeng Zhang, Zhenguo Li, Yi Zhong†, Jun Zhu†
    International Conference on Learning Representations (ICLR), 2022
    [Paper]
  • Analytic-DPM: an Analytic Estimate of the Optimal Reverse Variance in Diffusion Probabilistic Models
    Fan Bao, Chongxuan Li†, Jun Zhu†, Bo Zhang
    International Conference on Learning Representations (ICLR), 2022
    Outstanding Paper Award
    [Paper] [Code]
  • Probabilistic Neural-Symbolic Models with Inductive Posterior Constraints
    Ke Su, Hang Su, Chongxuan Li, Jun Zhu, Bo Zhang
    IEEE Transactions on Neural Networks and Learning Systems (TNNLS), 2022
  • Deep reinforcement learning with credit assignment for combinatorial optimization
    Dong Yan, Jiayi Weng, Shiyu Huang, Chongxuan Li, Yichi Zhou, Hang Su, Jun Zhu
    Pattern Recognition, 2022 (Runner-up Prize of IEEE VIZDoom RL Competition 2017)
    [Paper]
    • 2021
  • Triple Generative Adversarial Networks
    Chongxuan Li, Kun Xu, Jun Zhu, Jiashuo Liu, Bo Zhang.
    IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2021
    [Paper] [Code]
  • Stability and Generalization of Bilevel Programming in Hyperparameter Optimization
    Fan Bao*, Guoqiang Wu*, Chongxuan Li*, Jun Zhu, Bo Zhang
    Advances in Neural Information Processing Systems (NeurIPS), 2021
    [Paper]
  • Rethinking and Reweighting the Univariate Losses for Multi-Label Ranking: Consistency and Generalization
    Guoqiang Wu*, Chongxuan Li*, Kun Xu, Jun Zhu
    Advances in Neural Information Processing Systems (NeurIPS), 2021
    [Paper]
  • ORDisCo: Effective and Efficient Usage of Incremental Unlabeled Data for Semi-supervised Continual Learning
    Liyuan Wang, Kuo Yang, Chongxuan Li†, Lanqing Hong, Zhenguo Li, Jun Zhu†
    The IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR), 2021
    [Paper]
  • Variational (Gradient) Estimate of the Score Function in Energy-based Latent Variable Models
    Fan Bao, Kun Xu, Chongxuan Li, Lanqing Hong, Jun Zhu, Bo Zhang
    International Conference on Machine Learning (ICML), 2021
    [Paper] [Code]
  • Implicit Normalizing Flows
    Cheng Lu, Jianfei Chen, Chongxuan Li, Qiuhao Wang, Jun Zhu
    International Conference on Learning Representations (ICLR), 2021
    Spotlight Presentation
    [Paper] [Code]
  • MiCE: Mixture of Contrastive Experts for Unsupervised Image Clustering
    Tsung Wei Tsai, Chongxuan Li, Jun Zhu
    International Conference on Learning Representations (ICLR), 2021
    [Paper] [Code]
    • 2020
  • Efficient Learning of Generative Models via Finite-Difference Score Matching
    Tianyu Pang*, Kun Xu*, Chongxuan Li, Yang Song, Stefano Ermon, Jun Zhu
    Advances in Neural Information Processing Systems (NeurIPS), 2020
    [Paper] [Code]
  • Bi-level Score Matching for Learning Energy-based Latent Variable Models
    Fan Bao*, Chongxuan Li*, Hang Su, Jun Zhu, Bo Zhang
    Advances in Neural Information Processing Systems (NeurIPS), 2020
    [Paper] [Code]
  • Learning Implicit Generative Models by Teaching Density Estimators
    Kun Xu*, Chao Du*, Chongxuan Li, Jun Zhu, Bo Zhang
    European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML/PKDD), 2020
    [Paper] [Code]
  • Understanding and Stabilizing GANs' Training Dynamics with Control Theory
    Kun Xu, Chongxuan Li, Huanshu Wei, Jun Zhu, Bo Zhang
    International Conference on Machine Learning (ICML), 2020
    [Paper] [Code]
  • To Relieve Your Headache of Training an MRF, Take AdVIL
    Chongxuan Li, Chao Du, Kun Xu, Max Welling, Jun Zhu, Bo Zhang
    International Conference on Learning Representations (ICLR) 2020
    [Paper]
    • 2019
  • Conditional Graphical Generative Adversarial Networks
    Chongxuan Li, Jun Zhu, Bo Zhang
    Journal of Software (in Chinese), 2019
  • Multi-objects Generation with Amortized Structural Regularization
    Kun Xu, Chongxuan Li, Jun Zhu and Bo Zhang
    Advances in Neural Information Processing Systems (NeurIPS) 2019
    [Paper]
    • 2018
  • Max-Margin Deep Generative Models for (Semi-)Supervised Learning
    Chongxuan Li, Jun Zhu, Bo Zhang.
    IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2018
    [Paper] [Code]
  • Graphical Generative Adversarial Networks
    Chongxuan Li, Max Welling, Jun Zhu and Bo Zhang
    Advances in Neural Information Processing Systems (NeurIPS) 2018
    [Paper] [Code]
  • Learning to Write Stylized Chinese Characters by Reading a Handful of Examples
    Danyang Sun, Tongzheng Ren, Chongxuan Li, Jun Zhu, Hang Su, Bo Zhang
    International Joint Conferences on Artificial Intelligence (IJCAI) 2018
    [Paper]
  • Collaborative Filtering with User-Item Co-Autoregressive Models
    Chao Du , Chongxuan Li, Yin Zheng, Jun Zhu, Bo Zhang
    Thirty-Second AAAI Conference on Artificial Intelligence (AAAI), 2018
    [Paper] [Code]
    • 2017
  • Triple Generative Adversarial Nets
    Chongxuan Li, Kun Xu, Jun Zhu, Bo Zhang
    Advances in Neural Information Processing Systems (NeurIPS), 2017
    [Paper] [Code]
  • Population Matching Discrepancy and Applications in Deep Learning
    Jianfei Chen, Chongxuan Li, Yizhong Ru, Bo Zhang
    Advances in Neural Information Processing Systems (NeurIPS), 2017
    [Paper]
    • 2016
  • Towards Better Analysis of Deep Convolutional Neural Networks
    Mengchen Liu, Jiaxin Shi, Zhen Li, Chongxuan Li, Jun Zhu, Shixia Liu
    IEEE VAST 2016, TVCG track, 2016
    [Paper] [Demo]
  • Learning to Generate with Memory
    Chongxuan Li, Jun Zhu, Bo Zhang
    International Conference on Machine Learning (ICML) 2016
    [Paper] [Code]
    • 2015
  • Max-Margin Deep Generative Models
    Chongxuan Li, Jun Zhu, Tianlin Shi, Bo Zhang
    Advances in Neural Information Processing Systems (NeurIPS) 2015
    [Paper] [Code]

Selected Awards

Grant Support

  • Study of Efficient and Convergent Learning and Inference Algorithms in Deep Generative Models
    National Natural Science Foundation – General Program
  • Bayesian Deep Learning: Algorithms, Models and Applications
    Chinese Postdoctoral Innovative Talent Support Program

Talk

  • Diffusion Probabilistic Models: Foundations, Fast Inference and Controllable Generation
    Forum of Techniques in Foundation Model, Beijing Academy of Artificial Intelligence BAAI, online, 2022-12-17
    [Slides] [Blog]
  • Diffusion Probabilistic Models and Fast Inference Algorithms by Estimating the Optimal Reverse Variance
    VALSE Webinar, online, 2022-08-03
    [Slides]
  • Stability and Generalization of (Gradient-Based) Bilevel Programming in Hyperparameter Optimization
    Beijing Academy of Artificial Intelligence, BAAI-Live, online, 2021
    [Slides]
  • Deep Generative Models: Representation, Learning and Inference
    Tutorial of unsupervised learning in Vision and Learning Seminar (VALSE) @ Hangzhou, 2021
    A short version is also presented in MLNLP community, online, 2021
    [Slides]
  • Learning Generative Adversarial Networks with Limited Supervision
    GAN Workshop in National Big Data and Artificial Intelligence Science Conference (CSIAM-BDAI) @ Kunming, 2019
  • Max Margin Deep Generative Models for (Semi)-supervised Learning
    Student Webinar in Vision and Learning Seminar (VALSE) @ Hefei, 2019

Book Chapter

  • 可解释人工智能导论（第二章 贝叶斯方法），电子工业出版社，2022年4月

Service

  • Reviewer or (Senior) PC member: ICML, NeurIPS, ICLR, UAI, TPAMI, TCYB, TNNLS