Rongzhen Wang (王榕甄)
Email: wangrz@ruc.edu.cn |
|
Biography
I am a third-year PhD student in the GSAI-ML Group, Renmin University of China, advised by Prof. Chongxuan Li. Prior to PhD, I received my Bachelor's degree from the School of Statistics at Renmin University of China in 2023.
I am interested in optimization and architectural improvements for large-scale model training, particularly through theoretical tools and systematic experiments.
Previously, I have worked on the theoretical analysis and method design of optimization algorithms and generative models.
Selected Publications
* indicates the equal contribution-
LLaDA 1.5: Variance-Reduced Preference Optimization for Large Language Diffusion Models
Fengqi Zhu*, Rongzhen Wang*, Shen Nie, Xiaolu Zhang, Chunwei Wu, Jun Hu, Jun Zhou, Jianfei Chen, Yankai Lin, Ji-Rong Wen, Chongxuan Li
Annual Meeting of the Association for Computational Linguistics (ACL), 2026
[Code] [Model] -
A Theory for Conditional Generative Modeling on Multiple Data Sources
Rongzhen Wang, Yan Zhang, Chenyu Zheng, Chongxuan Li, Guoqiang Wu
International Conference on Machine Learning (ICML), 2025
[Code] [Slides] -
Lower Bounds of Uniform Stability in Gradient-Based Bilevel Algorithms for Hyperparameter Optimization
Rongzhen Wang, Chenyu Zheng, Guoqiang Wu, Xu Min, Xiaolu Zhang, Jun Zhou, Chongxuan Li
Advances in Neural Information Processing Systems (NeurIPS), 2024
[Slides]
Other Publications
-
Scaling Diffusion Transformers Efficiently via μP
Chenyu Zheng, Xinyu Zhang, Rongzhen Wang, Wei Huang, Zhi Tian, Weilin Huang, Jun Zhu, Chongxuan Li
Advances in Neural Information Processing Systems (NeurIPS), 2025
[Code] [机器之心] -
ControlVideo: Conditional Control for One-shot Text-driven Video Editing and Beyond
Min Zhao, Rongzhen Wang, Fan Bao, Chongxuan Li, Jun Zhu
Science China Information Sciences (SCIS), 2025
[Code] [Project Page] -
On Mesa-Optimization in Autoregressively Trained Transformers: Emergence and Capability
Chenyu Zheng, Wei Huang, Rongzhen Wang, Guoqiang Wu, Jun Zhu, Chongxuan Li
Advances in Neural Information Processing Systems (NeurIPS), 2024
[Code] [Slides]
Learning & Sharing
I like to take notes :) Here are some of my presentations and notes.- Tutorial on group session, Basics of Reinforcement Learning and Its Applications in LLM Alignment, 7.2025 [Slides in Chinese]
- Tutorial on group session, A Brief Introduction to Learning Theory, 11.2024 [Slides]
- Report on group session, Algorithmic Stability of Bilevel (Stochastic) Gradient Descent, 3.2024 [Slides]
- Report on group session, Algorithmic Stability of (Stochastic) Gradient Descent, 12.2023 [Slides]