Xianbang Wang

I am an undergraduate at MIT, studying mathematics with computer science and graduating in May 2027. I do research with Prof. Kaiming He.

I build and train generative models from scratch. I am a huge believer in Occam’s razor and the bitter lesson, and to me research is about subtraction: taking parts away until only what matters is left.

I want to do research that people actually use. I believe AGI is coming and that it will do enormous good for humanity, and I would like to help build it well.

My motto is keep the beginner’s mind. Few things feel better than stepping into a field I know nothing about. Lately that means organizing a group of AI agents to work together the way a team of human experts does.

Before MIT, I grew up on math and programming olympiads (IMO 2024 Gold Medal).

Things I love: rhythm games (Phigros, maimai, Arcaea), Manchester United, school anime, Chinese and Japanese pop, scheduling systems, teaching, eating, traveling with friends, and solving math and programming olympiad problems without pen or paper (purely for fun).

Things I hate: two-faced people, jealousy, rhinitis, and the finite context window of humans.

CV · Email · GitHub · X · Google Scholar · LinkedIn

Blogs

MiniT2I: A Minimalist Baseline for Text-to-Image Generation [gallery] X. Wang, H. Zhao, Y. Lu, K. Zhou, L. Ma, K. He June 2026

Publications

A Better Critic Matters for Distribution Matching Distillation [project page] Z. Wu*, X. Wang*, Z. Geng, M. Cao, Z. Zhang, R. Zhang, E. Shechtman, R. Cao Submitted to ICLR 2027
Bidirectional Normalizing Flow: From Data to Noise and Back Y. Lu*, Q. Sun*, X. Wang*, Z. Jiang, H. Zhao, K. He CVPR 2026, Highlight
One-step Latent-free Image Generation with Pixel Mean Flows Y. Lu*, S. Lu*, Q. Sun*, H. Zhao*, Z. Jiang, X. Wang, T. Li, Z. Geng, K. He ICML 2026
SANA-Video: Efficient Video Generation with Block Linear Diffusion Transformer Junsong Chen*, Yuyang Zhao*, Jincheng Yu*, R. Chu, Junyu Chen, S. Yang, X. Wang, Y. Pan, D. Zhou, H. Ling, H. Liu, H. Yi, H. Zhang, M. Li, Y. Chen, H. Cai, S. Fidler, P. Luo, S. Han, E. Xie ICLR 2026, Oral
LongLive: Real-time Interactive Long Video Generation S. Yang, W. Huang, R. Chu, Y. Xiao, Y. Zhao, X. Wang, M. Li, E. Xie, Yingcong Chen, Y. Lu, S. Han, Yukang Chen ICLR 2026
H3DP: Triply-Hierarchical Diffusion Policy for Visuomotor Learning Y. Lu*, Y. Tian*, Z. Yuan*, X. Wang, P. Hua, Z. Xue, H. Xu ICLR 2026

* Equal contribution.