Generative AI meets Physics

by Dr Kai Zhou (CUHK-Shenzhen)

Monday 1 Dec 2025, 10:30 → 11:30 Asia/Shanghai

S408

In this talk I will present some of the applications of generative AI models in exploring physics, including those towards QCD matter under extreme conditions. Several modern generative models, including GAN, autoregressive model, normalizing flow model and diffusion model will be mentioned for their physics application, especially in the context of lattice field theory configuration generation. The connection between diffusion model and stochastic quantization in the context of lattice field theory will be explained. I will also introduce higher order cumulants evaluation on these diffusion models. In the end some applications of diffusion model in the form of point cloud to construct fast emulator for heavy ion collision simulation will be talked about, which may promise the Bayesian inference for physics from these complicated physical processes.

Bio:
Dr. Kai Zhou received his B.Sc. degree in Physics from Xi'an Jiaotong University in 2009, and his PhD degree in Physics from Tsinghua University in 2014. After that, he worked as a Postdoctoral researcher in the Institute for Theoretical Physics (ITP) at Goethe University Frankfurt in Germany from 2014 to 2017. Since 2017 he started as a Research Fellow (W1 professor status) Group Leader at the Frankfurt Institute for Advanced Studies (FIAS), leading the AI for Science group “Deepthinkers”,  focusing on physics studies with modern computational paradigms Machine- and Deep-Learning, supervises Master/PhD students and Postdocs in AI for Science. From 2022 he was then promoted to Fellow (W2 status) at FIAS. He joined CUHK-Shenzhen as an Assistant Professor since the end of 2023.