Portrait of Meng Xu outdoors

Meng Xu

Ph.D. candidate in Computational Mathematics

Academy of Mathematics and Systems Science
Chinese Academy of Sciences

About Me

I am a Ph.D. candidate in Computational Mathematics at the Academy of Mathematics and Systems Science, Chinese Academy of Sciences, advised by Prof. Ya-Feng Liu. I'm also very fortunate to work with Prof. Anthony Man-Cho So at The Chinese University of Hong Kong and Prof. Bo Jiang at Nanjing Normal University. Previously, I received my B.Sc. in Computational Mathematics from the School of Mathematics, Jilin University.

My research focuses on mathematical optimization, particularly Riemannian nonsmooth and minimax optimization, and its applications in machine learning. I am currently especially interested in LLM post-training and LLM agents. I also spent the summer of 2026 at Alibaba Group, working on LLM post-training for creative writing.

Education

09/2022 – 06/2027 (expected)Ph.D. candidate, AMSS, Chinese Academy of Sciences. Advisor: Prof. Ya-Feng Liu.

09/2018 – 06/2022B.Sc. in Computational Mathematics, School of Mathematics, Jilin University.

Experience

06/2026 – 09/2026

AI Algorithm Intern · Alibaba Group

LLM post-training for creative writing

Publications

Google Scholar
  1. An inexact proximal framework for nonsmooth Riemannian difference-of-convex optimization

    B. Jiang, M. Xu, X. Cai, and Y.-F. Liu

    SIAM Journal on Optimization2026AcceptedarXiv

  2. A Riemannian alternating descent ascent algorithmic framework for nonconvex-linear minimax problems on Riemannian manifolds

    M. Xu, B. Jiang, Y.-F. Liu, and A. M.-C. So

    Mathematics of Operations Research2026AcceptedarXiv

  3. On the oracle complexity of a Riemannian inexact augmented Lagrangian method for nonsmooth composite problems over Riemannian submanifolds

    M. Xu, B. Jiang, Y.-F. Liu, and A. M.-C. So

    Optimization Letters2025AcceptedarXiv

  4. An efficient alternating Riemannian/projected gradient descent ascent algorithm for fair principal component analysis

    M. Xu, B. Jiang, W. Pu, Y.-F. Liu, and A. M.-C. So

    IEEE ICASSP2024pp. 7195–7199arXiv

  5. A stochastic Riemannian alternating descent ascent method for nonsmooth composite expectation optimization on Riemannian manifolds

    M. Xu, B. Jiang, Y.-F. Liu, and A. M.-C. So, 2026. arXiv

  6. Riemannian difference-of-convex optimization for K-means clustering

    M. Xu, B. Jiang, H. Zhang, Y.-F. Liu, and A. M.-C. So, 2026. arXiv

  7. NextEval: An agent-based solver and benchmark for derivative-free optimization

    S. Chen, L. Cui, C. Huang, Y.-F. Liu, M. Xie, M. Xu, and Z. Zhang, 2026. arXiv website

Honors

  • 2026, 2021, 2020National ScholarshipChinese Academy of Sciences (2026); Jilin University (2020, 2021)
  • 2024Hua Loo-Keng ScholarshipAMSS, Chinese Academy of Sciences
  • 2022Outstanding Graduate & Wang Xianghao ScholarshipJilin University
  • 2020Meritorious WinnerMathematical Contest in Modeling, COMAP

Selected Talks

Jul 2025

A Flexible Algorithmic Framework for Nonconvex-Linear Minimax Problems on Riemannian Manifolds

ICCOPT · Los Angeles

Journal Reviewer

  • SIAM Journal on Optimization
  • IEEE Transactions on Signal Processing