Leitian Tao

I am a Ph.D. student at University of Wisconsin–Madison in the Computer Sciences department. My current research interests focus on large language model alignment, reasoning, and synthetic data generation.

I work as a research assistant at the University of Wisconsin–Madison, under the supervision of Sharon Li.

I am currently working as a Research Scientist Intern at Meta FAIR, focusing on Large Language Models and alignment research. Previously, I worked as a summer research intern at Adobe Research, focusing on Large Language Models for code generation.

I graduated from Wuhan University (WHU), where my advisor was Zhenzhong Chen. I have also worked as a research intern at Shanghai AI Lab, where my mentor was Peng Gao.

Email  /  Google Scholar  /  Github /  LinkedIn /  CV / 

My research interests span machine learning, computer vision, and natural language processing. I focus particularly on large language model alignment, reasoning with synthetic data, and open-world machine learning. Recent work shows that synthetic data from smaller, weaker models can be more effective than data from larger models for improving LLM reasoning. Representative papers are highlighted. (* indicates equal contribution)

Journal Reviewer: IJCV

Conference Reviewer: ICML, ICLR, NeurIPS

Template inspired by Jon Barron's website.
Last updated: June 2025