Zhehang Du

I am a PhD student in the Department of Statistics and Data Science at The Wharton School of the University of Pennsylvania, advised by Prof. Weijie Su. Previously, I received a BS in Mathematics and Applied Mathematics from the University of Science and Technology of China and interned at Microsoft Research Asia under the supervision of Huishuai Zhang.

My research focuses on optimization and the mathematical foundations of modern AI systems. I am interested in developing principled methods for understanding, training, and improving large-scale learning systems, with a growing interest in new architectures such as world models.

Selected Papers

My research publications are listed on Google Scholar.

• Zhehang Du, Weijie Su. Polar Switch: Fast Muon Orthogonalization with Polynomial Circuits. arXiv preprint arXiv:2610.26742, 2026. [Code]
• Zhehang Du, Weijie Su. The Newton-Muon Optimizer. arXiv preprint arXiv:2604.01472, 2026. [Code] [Blog]
• Zhehang Du, Hangfeng He, Weijie Su. Uncovering Symmetry Transfer in Large Language Models via Layer-Peeled Optimization. arXiv preprint arXiv:2605.12756, 2026.
• Zihan Zhu, Zhehang Du, Xuyang Chen, Tim Tsz-Kit Lau, Jiayuan Wu, X. Y. Han, Qi Long, Weijie Su. Rank-Efficient LoRA via Joint Tangent-Space Optimization under Isotropic Curvature. arXiv preprint arXiv:2609.12123, 2026.
• Humanity’s Last Exam. Nature, 2026. [Code]
• Agents’ Last Exam. In The Fortieth Annual Conference on Neural Information Processing Systems Evaluations and Datasets Track, 2026. [Code]