Passion and Thinking are always my best teachers

I'm a PhD student at UW-Madison, where I'm very honored to be advised by Prof. Stephen Wright (2024.09 - now) and co-advised by Prof. Jiawei Zhang (2025.07 - now)! Previously, I had the privilege of working with Prof. Qin Li (2024.09 - 2025.05) and Prof. Yang Yuan (2023.04 - 2024.03), from whom I learned a lot!

I have a keen interest in exploring how the Universe and Intelligence work and how can we possibly understand them. In this era, AI is dependent on data-driven deep learning methods, which is unexplainable and uncontrollable. I believe that theory-guided intelligence is the future of AI. My long-term goal is to embrace the superintelligence via fully explainable methods! 🌞

We know how it works, what it learns, and why it excels!

Another noteworthy thing is that I'm a devoted advocate of Slow Science, or rather, a faithful of Thinking.

I'm always open to visiting and collaborating with leading academic and industrial research groups. If you see a potential fit or are interested in my research, please feel free to contact me! 

⭐ Recent News ⭐

02/2026. Glad that the first paper in my PhD stage is accepted by the CVPR2026!

08/2024. PhD journey at UW-Madison starts! Super glad to work with my mentors and friends on exciting topics!

Representative Papers

An Optimization Framework for Denoising Generation

Xinyu Zhou, Jiawei Zhang, Stephen Wright. (In Submission)

Smoothing the Score Function to Enhance Generalization in Diffusion Models

Xinyu Zhou, Jiawei Zhang, Stephen Wright. [PDF] 

In the 43rd Conference on Computer Vision and Pattern Recognition (CVPR), 2026.

Talk

IFDS Seminar - "Smoothing the Score Function to Enhance Generalization in Diffusion Models"; UW-Madison, 2026.02