About Me

Hi there 👋 I am Xiaowei Yin, a first year PhD student in Computer Science in UNC, supervised by Prof. Zhun Deng. Previously, I completed my undergraduate study at the School of Statistics, Renmin University of China (RUC), majoring in Statistics.

Research

My research interests lies on the intersection of AI and statistics. It’s both fun and painful to work between theory and empirics, but I’m sure this path will pay off — combining them usually leads to more insightful and creative ideas.

• AI safety

I am interested in the broad AI safety topic, with a paticular interest in using statistical methods to enhance LLM reasoning, makeing it more calibrated and reliable. I hope to develop frameworks that combine theoretical rigor with practical efficiency in enhancing AI safety.

• Learning theory (ML,RL,DL)

I find it interesting that learning theory can characterize the process of “learning” by leveraging concentration inequalities and several measures of complexity, including Rademacher Complexity, VC dimension, and Covering Numbers. Bridging them together is a satisfying intellectual pursuit! 😄 I hope to dig deeper into this field not only because I find it interesting but also because I believe insightful ideas can be born from this theoretical analysis into ML, RL, and DL. I hope to contribute meaningfully to this field.