Zheyang Xiong

PhD Student, Computer Science, University of Wisconsin-Madison

prof_pic.jpg

I am a PhD student in Computer Science at the University of Wisconsin-Madison, advised by Dimitris Papailiopoulos. I received my B.S. in Mathematics and B.A. in Computer Science from Rice University in 2023, advised by Anastasios Kyrillidis.

I study Large Language Models, with recent focus on efficient reasoning and scaling laws.

I am also a part-time student researcher at Microsoft Research AI Frontiers Lab.

publications

* denotes equal contribution.

  1. Preprint
    SuperThoughts: Reasoning Tokens in Superposition
    Zheyang Xiong, Shivam Garg*, Max Yu*, and 4 more authors
    2025
    Paper
  2. ICML
    Everything Everywhere All at Once: LLMs can In-Context Learn Multiple Tasks in Superposition
    Zheyang Xiong, Ziyang Cai, John Cooper, and 11 more authors
    The International Conference on Machine Learning (ICML), 2025 Spotlight
    Paper
  3. ICLR
    From Artificial Needles to Real Haystacks: Improving Retrieval Capabilities in LLMs by Finetuning on Synthetic Data
    Zheyang Xiong, Vasilis Papageorgiou, Kangwook Lee, and 1 more author
    International Conference on Learning Representations (ICLR), 2025
    Paper
  4. ICLR
    OmniEdit: Building Image Editing Generalist Models Through Specialist Supervision
    Cong Wei*, Zheyang Xiong*, Weiming Ren, and 3 more authors
    International Conference on Learning Representations (ICLR), 2025
    Paper
  5. ICML
    Can Mamba Learn How to Learn? A Comparative Study on In-Context Learning Tasks
    Jongho Park, Jaeseung Park, Zheyang Xiong, and 5 more authors
    International Conference on Machine Learning (ICML), 2024
    Paper
  6. AISTATS
    Strong Lottery Ticket Hypothesis with ε–perturbation
    Zheyang Xiong*, Fangshuo Liao*, and Anastasios Kyrillidis
    International Conference on Artificial Intelligence and Statistics (AISTATS), 2023
    Paper