About
I am a Ph.D. student at the University of Michigan, Ann Arbor, advised by Prof. Qing Qu. I received my B.E. from the University of Electronic Science and Technology of China.
My research explores how generative models learn, generalize, memorize data. I am also interested in multi-agent latent.
Explore my research projects, publications, and CV.
Selected Publications
View All →Generalization of Diffusion Models Arises with a Balanced Representation Space
Zekai Zhang, Xiao Li, Xiang Li, Lianghe Shi, Meng Wu, Molei Tao, Qing Qu
ICLR (Poster) · 2026
Connecting balanced and spiky representations to generalization and memorization, with applications to detection and image editing.
A Closer Look at Model Collapse: From a Generalization-to-Memorization Perspective
Lianghe Shi, Meng Wu, Huijie Zhang, Zekai Zhang, Molei Tao, Qing Qu
NeurIPS (Spotlight) · 2025
Explaining collapse as a transition from generalization to memorization, and studying entropy-based data selection for recursive training.