Zhun Deng

Zhun Deng is a tenure-track Assistant Professor at the Department of Computer Science, University of North Carolina at Chapel Hill. His research bridges machine learning, statistics, and theoretical computer science, focusing on rigorous frameworks for responsible AI systems. He previously held postdoctoral positions at Columbia University and completed his Ph.D. at Harvard's Theory of Computation group under Cynthia Dwork.

• Ph.D. in Computer Science, Harvard University (2022)
• B.Sc. in Mathematics, Chu Kochen Honors College, Zhejiang University

His research spans theoretical foundations of machine learning, including:

• Quantile-based risk control and conformal prediction
• Fairness guarantees in algorithmic decision-making
• Uncertainty quantification for LLMs
• Copyright frameworks for generative AI
• Physics-informed hybrid models
• Multi-agent reinforcement learning with constraints

Recent work analyzes LLM alignment through distribution-free methods (ICML 2025), explores performativity challenges (ICML 2025), and develops calibration techniques (ICLR 2024). Collaborations include research interns from Stanford, MIT, and NYU.

Students in his group include:

• Ruomeng Ding (Ph.D., UNC)
• Xiaowei Yin (Ph.D., UNC)
• Kaicheng Zhang (Ph.D., UNC)