I am an assistant professor in the School of Data Science and Society and Department of Statistics and Operations Research at UNC-Chapel Hill. I received my Ph.D. in Computer Science from Duke University.

My research focuses on understanding and controlling model diversity in modern machine learning, and on building trustworthy and reliable AI systems. A central theme of my work is the Rashomon Effect, the existence of many equally accurate yet behaviorally distinct models, and its implications for scientific understanding of learning and model behavior. Current key directions include:

1. characterize the set of near-optimal models or equally valid outputs, and study the variability that arises across them
2. leverage model diversity to reveal alternative explanations and failure modes and support human-aligned, interpretable, and stable reasoning in modern ML systems and LLMs.

Lab Openings: I am actively recruiting highly motivated PhD students and multiple research interns with strong technical backgrounds. Candidates who are interested are strongly encouraged to contact me via email, including a resume and transcripts.