Jianhao Ma

jianhao.jpg

Department of Industrial Engineering

Tsinghua University

Beijing, China

jianhao@tsinghua.edu.cn

I am a tenure-track Assistant Professor in the Department of Industrial Engineering at Tsinghua University. Previously, I was a postdoctoral researcher at the University of Pennsylvania, working with Professor Yuxin Chen. I received my Ph.D. in Industrial and Operations Engineering from the University of Michigan, where I was fortunate to be advised by Professor Salar Fattahi.

Research interests

My research explores both AI for optimization and optimization for AI. On the AI-for-optimization side, I study how large language models can help solve optimization-theory problems and construct rigorous proofs. On the optimization-for-AI side, I use optimization theory to design better pretraining algorithms for large language models.

Join my group

I welcome inquiries from prospective Ph.D. students for future admission and can recruit one Ph.D. student per year; the position for this year has already been filled. I can also host undergraduate research assistants, both in person and remotely. If you are interested, please contact me by email.

news

Aug 2026 New preprint: A lower bound for stepsize-based acceleration of gradient descent, with Yuxin Chen.
Jul 2026 I joined the Department of Industrial Engineering at Tsinghua University as a tenure-track Assistant Professor.

selected recent publications

  1. A lower bound for stepsize-based acceleration of gradient descent
    Jianhao Ma and Yuxin Chen
    arXiv preprint, 2026
  2. WSqD: A Horizon-Free Learning Rate Schedule for Large Model Training
    Jianhao Ma and Yuxin Chen
    arXiv preprint, 2026
  3. Preconditioning Benefits of Spectral Orthogonalization in Muon
    Jianhao Ma, Yu Huang, Yuejie Chi, and Yuxin Chen
    arXiv preprint, 2026
  4. Convergence of Gradient Descent with Small Initialization for Unregularized Matrix Completion
    Jianhao Ma and Salar Fattahi
    In Conference on Learning Theory, 2024