Xiangyan Liu

School of Computing, National University of Singapore

lxy.png

Hello! I’m Xiangyan, a second-year PhD student in Computer Science at National University of Singapore, where I’m fortunate to be advised by Prof. Michael Qizhe Shieh.

My research focuses on multimodal and agentic foundation models, with an interest in solutions that are simple, generalizable, and scalable.

I previously interned at Moonshot AI, where I worked on visual agents. Before that, I interned at Alibaba Tongyi and Shanghai AI Lab. I also had the opportunity to work with Prof. Tao Lin at Westlake University.

I have collaborated closely with Dr. Jinjie Ni, Dr. Chao Du and Dr. Longxu Dou at NUS and Sea AI Lab. Recently, I have been working closely with Dr. Tianyu Pang.

Feel free to reach out if you would like to chat or discuss any ideas! You can find my contact information at the bottom of this page.

selected works

  1. Tech Report
    Kimi K2.5: Visual Agentic Intelligence
    Kimi Team
    Technical Report, 2026
    Core contributor to the visual agent for front-end development.
    arXiv Website
  2. ICLR 2026
    MCPMark: A Benchmark for Stress-Testing Realistic and Comprehensive MCP Use
    Zijian Wu*, Xiangyan Liu*, Xinyuan Zhang*, Lingjun Chen*, Fanqing Meng*, Lingxiao Du*, Yiran Zhao*, Fanshi Zhang*, Yaoqi Ye, Jiawei Wang, and 5 more authors
    In International Conference on Learning Representations, 2026
    arXiv Code
  3. NeurIPS 2025
    NoisyRollout: Reinforcing Visual Reasoning with Data Augmentation
    Xiangyan Liu*, Jinjie Ni*, Zijian Wu*, Chao Du, Longxu Dou, Haonan Wang, Tianyu Pang, and Michael Qizhe Shieh
    In Advances in Neural Information Processing Systems, 2025
    arXiv Code
  4. NAACL 2025
    CodexGraph: Bridging Large Language Models and Code Repositories via Code Graph Databases
    Xiangyan Liu*, Bo Lan*, Zhiyuan Hu, Yang Liu, Zhicheng Zhang, Fei Wang, Michael Shieh, and Wenmeng Zhou
    In Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics, 2025
    arXiv Code
  5. ICLR 2024
    Towards Robust Multi-Modal Reasoning via Model Selection
    Xiangyan Liu*, Rongxue Li*, Wei Ji, and Tao Lin
    In International Conference on Learning Representations, 2024
    arXiv Code
  6. MM 2023 (Oral)
    Online Distillation-Enhanced Multi-Modal Transformer for Sequential Recommendation
    Wei Ji*, Xiangyan Liu*, An Zhang, Yinwei Wei, Yongxin Ni, and Xiang Wang
    In ACM International Conference on Multimedia, 2023
    arXiv Code