Qunhong Zeng

Research Engineer at MiniMax

I work on post-training LLM for agentic coding.

I received my M.Eng. (2026) and B.Eng. (2023) from the School of Computer Science at Beijing Institute of Technology, where I was advised by Prof. Yuxia Zhang and worked on AI for Software Engineering (AI4SE).

Please feel free to contact me if you’re interested in having a discussion :)

Publications

  1. Evaluating Generated Commit Messages with Large Language Models

    Qunhong Zeng, Yuxia Zhang, Zexiong Ma, Bo Jiang, Ningyuan Sun, and 3 more authors
    ICSE 2026
  2. Enhancing Issue Localization Agent with Tool-Interactive Training

    Zexiong Ma, Chao Peng, Qunhong Zeng, Pengfei Gao, Yanzhen Zou, and 1 more author
    ICSE 2026
  3. SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Tasks

    Xiangyi Li, Yimin Liu, Wenbo Chen, Bingran You, …, Qunhong Zeng, and 73 more authors
    NeurIPS 2026
  4. The MiniMax-M2 Series: Mini Activations Unleashing Max Real-World Intelligence

    Aili Chen, Aonian Li, Baichuan Zhou, Bangwei Gong, …, Qunhong Zeng, and 213 more authors
    Technical Report, 2026
  5. OctoBench: Benchmarking Scaffold-Aware Instruction Following in Repository-Grounded Agentic Coding

    Deming Ding, Shichun Liu, Enhui Yang, Jiahang Lin, …, Qunhong Zeng, and 10 more authors
    ACL 2026
  6. A First Look at Conventional Commits Classification

    Qunhong Zeng, Yuxia Zhang, Zhiqing Qiu, and Hui Liu
    ICSE 2025

Experience

  1. MiniMax Research Engineer

    Jul 2025 – Present
    • Work on data infrastructure and SFT/RL data pipelines for coding post-training.
    • Contributed to the agentic coding capabilities of the MiniMax-M2 and MiniMax-M3 series.
  2. Meituan Beidou Program Intern

    Jun 2025 – Jul 2025
    • Researched multi-turn conversational RL to enhance LLM’s dialogue capabilities.
  3. ByteDance Research Intern, Trae

    Aug 2024 – Jun 2025
    • Worked on LLM post-training for Trae, including code editing and issue localization.
    • Contributed to the open-source RL framework verl.
  4. Oneflow Framework Intern

    Mar 2024 – Sep 2024
    • Added Ascend NPU support to the OneFlow deep learning framework.
  5. Momenta HD-Map Backend Intern

    Sep 2022 – Jan 2023
    • Built Python/C++ extensions for the HD-map library, with a 20x speedup.
  6. ByteDance AI-Lab Intern

    Dec 2021 – Jun 2022
    • Worked on ByteDance’s machine learning platform.