Chenguang Wang

I am a PhD student in Computer Science at The Chinese University of Hong Kong, Shenzhen, under the supervision of Prof. Tianshu Yu.

Prior to this, I obtained my Master’s degree from the University of Chinese Academy of Sciences under the supervision of Prof. Tiande Guo, and my Bachelor’s degree in Mathematics from Zhengzhou University.

Research Interests

My research interests lie at the intersection of machine learning and probabilistic inference, with a particular focus on:

Neural Sampling: Developing efficient sampling algorithms for complex discrete and continuous distributions
Neural Combinatorial Optimization: Designing learning-based solvers with emphasis on generalization and scalability
AI4Science: Applying AI techniques to scientific discovery and problem-solving
Diffusion LLM: Exploring diffusion models for large language model training and inference

For a complete list of my publications, please visit my Google Scholar profile.

Experience

Research Intern

Shanghai Artificial Intelligence Laboratory

2025.07 - Present

Working on flow matching-based chemical retrosynthesis generation algorithms

Research Intern

Peking University & King’s College London

2021.11 - 2022.06

Game-theoretic approaches for combinatorial optimization
Spotlight presentation at ICLR 2022 Workshop

I am always open to research collaborations and discussions. Feel free to reach out via email!

Selected Publications

ICML
Sampling from Binary Quadratic Distributions via Stochastic Localization

Chenguang Wang, Kaiyuan Cui, Weichen Zhao, and 1 more author

arXiv preprint arXiv:2505.19438, May 2025

Abs arXiv Bib

We propose an efficient discrete sampling algorithm based on stochastic localization, achieving consistent performance improvements across multiple standard sampling benchmarks.
@article{wang2025sampling, title = {Sampling from Binary Quadratic Distributions via Stochastic Localization}, author = {Wang, Chenguang and Cui, Kaiyuan and Zhao, Weichen and Yu, Tianshu}, journal = {arXiv preprint arXiv:2505.19438}, year = {2025}, month = may, }

KDD

Understanding Oversmoothing in Diffusion-Based GNNs From the Perspective of Operator Semigroup Theory

Weichen Zhao, Chenguang Wang, Xinyan Wang, and 3 more authors

In Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V. 1, Jan 2025

Bib

@inproceedings{zhao2025understanding,
  title = {Understanding Oversmoothing in Diffusion-Based GNNs From the Perspective of Operator Semigroup Theory},
  author = {Zhao, Weichen and Wang, Chenguang and Wang, Xinyan and Han, Congying and Guo, Tiande and Yu, Tianshu},
  booktitle = {Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V. 1},
  pages = {2043--2054},
  year = {2025},
  month = jan,
}

TMLR

Efficient Training of Multi-task Neural Solver for Combinatorial Optimization

Chenguang Wang, Zhang-Hua Fu, Pinyan Lu, and 1 more author

Transactions on Machine Learning Research, Mar 2025

Bib

@article{wang2025efficient,
  title = {Efficient Training of Multi-task Neural Solver for Combinatorial Optimization},
  author = {Wang, Chenguang and Fu, Zhang-Hua and Lu, Pinyan and Yu, Tianshu},
  journal = {Transactions on Machine Learning Research},
  year = {2025},
  month = mar,
}

TPAMI

ASP: Learn a Universal Neural Solver!

Chenguang Wang, Zhouliang Yu, Stephen McAleer, and 2 more authors

IEEE Transactions on Pattern Analysis and Machine Intelligence, Nov 2024

Bib HTML

@article{wang2024asp,
  title = {ASP: Learn a Universal Neural Solver!},
  author = {Wang, Chenguang and Yu, Zhouliang and McAleer, Stephen and Yu, Tianshu and Yang, Yaodong},
  journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
  volume = {46},
  number = {6},
  pages = {4102--4114},
  year = {2024},
  month = nov,
  publisher = {IEEE},
}

Research Interests

Recent News

Experience

Selected Publications