Luhan Tang

My research focuses on diffusion models and large language models. I actively explore both theoretical foundations and empirical behavior in generative modeling, with specific research interests in:

  • Diffusion models: diffusion models for images and diffusion language models.
  • Large language models: multimodal systems and agentic models for complex tasks.
Portrait of Luhan Tang
Email / Google Scholar / GitHub / LinkedIn

I am a Ph.D. student at the University of California, Riverside, working with fantastic professors Greg Ver Steeg and Weixin Yao.

I received my B.S. degree in Statistics from Beijing Normal University, with minor coursework in Economics at Peking University.

Beyond research, I also enjoy photography and keeping a personal visual archive of places, light, and quiet everyday moments. Welcome to my photo gallery 📷

News

May '26
One paper accepted by ICML 2026!

Publications and Preprints

Filter by: show selected / show all by date / show Diffusion Language Models

Figure for sampler-centric diffusion preprint
Diffusion language model

Is Your Diffusion Sampler Actually Correct? A Sampler-Centric Evaluation of Discrete Diffusion Language Models

Luhan Tang, Longxuan Yu, Shaorong Zhang, Greg Ver Steeg

ICML 2026

Paper / Code

Teaching

Teaching Assistant, University of California, Riverside