Hey, I am Zach! I am a second-year CS PhD student at University of Chicago, advised by Ce Zhang.

My research interests are centered on large language models with a recent focus on self-evolving agents, reinforcement learning, test-time scaling, and interdisciplinary scientific applications.

I am Assistant Editor of the journal Data-centric Machine Learning Research (DMLR) under the JMLR family. I also contributed to Codabench, which is one of the most commonly used open source competition platforms according to ML Contests.

Zach Xu
If you are interested in collaborating on research, please feel free to reach out!

News

  • May2026Start research internship at Meta Superintelligence Labs
  • Apr2026Prefill Prefill-capable Decode (PPD) Disaggregation is accepted at ICML 2026
  • Jan2026Divide and Conquer Agents is accepted at ICLR 2026
  • Nov2025Start research internship at Together AI

Selected Publications

  • When Does Divide and Conquer Work for LLM in Processing Long Context? A Noise Decomposition Framework
    Zhen Xu, Shang Zhu, Jue Wang, Junlin Wang, Ben Athiwaratkun, Chi Wang, James Zou, Ce Zhang
    ICLR 2026[PDF]
    Invited presentations at Citadel Securities
  • Codabench: Flexible, easy-to-use, and reproducible meta-benchmark platform
    Zhen Xu, Sergio Escalera, Adrien Pavão, Magali Richard, Wei-Wei Tu, Quanming Yao, Huan Zhao, Isabelle Guyon
    Patterns Cell Press 2022[PDF]
    One of the most commonly used competition platforms according to ML Contests [2023, 2024, 2025]