Yuancheng Xu

PhD Student in Applied Mathematics
Unversity of Maryland, College Park
Email: ycxu [at] umd [dot] edu
About me

My name is Yuancheng Xu (徐源诚) and I am a PhD student at University of Maryland, College Park advised by Prof. Furong Huang. My research focuses on Trustworthy Machine Learning, including adversarial robustness, fairness and privacy of AI systems. I am particularly interested in automatic search for undesirable behaviors in (multimodal) large language models. I am also interested in understanding and enriching the reasoning and planning capabilities of machine learning systems.

Before joining UMD, I obtained the bachelor's degree in mathematics from Southern University of Science and Technology in China. I had worked with Prof. Christina Ramirez on developing machine learning algorithms for high dimensional longitudinal data with correlated feature at UCLA, and worked with Prof. Sukbin Lim on computational Neuroscience at NYU Shanghai.

I will be joining J.P. Morgan AI Research as a research intern in summer 2024.

I am always open to new ideas and collaborations. Feel free to drop me an email if you would like to brainstorm or collaborate with me.


[Exploring and Exploiting Decision Boundary Dynamics for Adversarial Robustness]
Exploring and Exploiting Decision Boundary Dynamics for Adversarial Robustness
Yuancheng Xu, Yanchao Sun, Micah Goldblum, Tom Goldstein, Furong Huang
International Conference on Learning Representations (ICLR), 2023
  • Mathematical characterization of how decison boundary moves.
  • Analytic expression for the margin's gradient. Directly applicable to robust training.
[Paper] [Code]