I’m Shukai Liu (刘书恺), also known as Shark Liu, a Ph.D. student at Beihang University, advised by Prof. Bo Jiang and Prof. Jian Yang.

My research focuses on code large language models and software engineering (SWE) agents, spanning scalable SWE data, long-horizon agent training, and multilingual code evaluation. At IQuest Research, I work on SWE evaluation, data and post-training as a core contributor to IQuest-Coder-V1 and IQuest-Q1.

Feel free to reach out via email for collaboration or discussion.

Selected Research

  • SWE-Crafter · NeurIPS 2026
    Scaling executable multilingual SWE data with meta-skill agents.

  • Context as a Tool (CaT) · Findings of ACL 2026
    Context management for long-horizon SWE agents.
    [paper] [arXiv]

  • McEval & MdEval · ICLR 2025 / Findings of ACL 2026
    Massively multilingual benchmarks for code generation and debugging.
    [McEval] [code] [MdEval]

🔥 News

  • 2026.09:  🎉 SWE-Crafter is accepted to NeurIPS 2026!
  • 2026.09:  SWE-Slim passes the Phase 1 review of AAAI 2027.
  • 2026.05:  🎉 Context as a Tool and MdEval are accepted to Findings of ACL 2026.
  • 2026.05:  🎉 From Diagrams to Code is accepted to ICML 2026.
  • 2026.03:  We release the IQuest-Coder-V1 technical report.
  • 2025.09:  I join IQuest Research and transfer to the Ph.D. program at Beihang University.
  • 2025.05:  🎉 M2RC-Eval is accepted to ACL 2025.
  • 2025.01:  🎉 McEval is accepted to ICLR 2025.

📝 Publications

* indicates equal contribution.

First-Author Papers

  • SWE-Crafter: Scaling Executable Multilingual Software Engineering Data with Meta-Skill Agents
    Shukai Liu, et al.
    NeurIPS 2026

  • Context as a Tool: Context Management for Long-Horizon SWE-Agents
    Shukai Liu, Bo Jiang, Jian Yang, Yizhi Li, Jinyang Guo, Xianglong Liu, Bryan Dai
    Findings of ACL 2026
    [paper] [arXiv]

  • SWE-Slim: Not All Successful Trajectories Are Worth Learning From
    Shukai Liu, et al.
    Under review at AAAI 2027

  • HINT: Learning the Harness Protocol for Interface-Invariant SWE Agents
    Shukai Liu, et al.
    Under review at ICLR 2027

  • MdEval: Massively Multilingual Code Debugging
    Shukai Liu*, Linzheng Chai*, Jian Yang*, Jiajun Shi, He Zhu, Liran Wang, Ke Jin, Wei Zhang, Hualei Zhu, Shuyue Guo, Tao Sun, Jiaheng Liu, Yunlong Duan, Yu Hao, Liqun Yang, Guanglin Niu, Ge Zhang, Zhoujun Li
    Findings of ACL 2026
    [paper] [arXiv]

  • McEval: Massively Multilingual Code Evaluation
    Linzheng Chai*, Shukai Liu*, Jian Yang*, Yuwei Yin, Ke Jin, Jiaheng Liu, Tao Sun, Ge Zhang, Changyu Ren, Hongcheng Guo, Zekun Wang, Boyang Wang, Xianjie Wu, Bing Wang, Tongliang Li, Liqun Yang, Sufeng Duan, Zhoujun Li
    ICLR 2025
    [paper] [code] [project]

Co-Authored Papers

  • From Diagrams to Code: Multilingual Programming with Visual Design
    Linzheng Chai, Jian Yang, Shukai Liu, Wei Zhang, Liran Wang, Ke Jin, Tao Sun, Congnan Liu, Chenchen Zhang, Hualei Zhu, Jiaheng Liu, Xianjie Wu, Ge Zhang, Tianyu Liu, Zhoujun Li
    ICML 2026
    [paper] [arXiv]

  • M2RC-Eval: Massively Multilingual Repository-level Code Completion Evaluation
    Jiaheng Liu*, Ken Deng*, Congnan Liu, Jian Yang, Shukai Liu, He Zhu, Peng Zhao, Linzheng Chai, Yanan Wu, Ke Jin, Ge Zhang, Zekun Wang, Guoan Zhang, Bangyu Xiang, Wenbo Su, Bo Zheng
    ACL 2025
    [paper] [arXiv]

  • LoopCoder: Scaling Code Intelligence via Looped Language Models
    Jian Yang, Wei Zhang, Shuyue Guo, Yizhi Li, Linzheng Chai, Zhengmao Ye, Shukai Liu, et al.
    Findings of ACL 2026
    [paper]

  • MIRA: Mid-training Rubric Anchoring for Source-Aware Data Selection
    Haowen Wang, Yaxin Du, Jian Yang, Jiajun Wu, Shukai Liu, Yuxuan Zhang, Pingjie Wang, Siheng Chen, Tuney Zheng, Ming Zhou, Xianglong Liu, Bryan Dai
    arXiv 2026
    [paper]

  • IQuest-Coder-V1 Technical Report
    Jian Yang, Wei Zhang, Shawn Guo, Zhengmao Ye, Lin Jing, Shark Liu, et al. (core contributor)
    arXiv 2026
    [paper] [model]

  • From Code Foundation Models to Agents and Applications: A Comprehensive Survey and Practical Guide to Code Intelligence
    Jian Yang, Xianglong Liu, Weifeng Lv, Ken Deng, Shawn Guo, Lin Jing, Yizhi Li, Shark Liu, et al.
    arXiv 2025
    [paper]

  • IFEvalCode: Controlled Code Generation
    Jian Yang, Wei Zhang, Shukai Liu, Linzheng Chai, Yingshui Tan, Jiaheng Liu, Ge Zhang, Wangchunshu Zhou, Guanglin Niu, Zhoujun Li, Binyuan Hui, Junyang Lin
    arXiv 2025
    [paper] [project]

  • FullStack Bench: Evaluating LLMs as Full Stack Coders
    ByteDance Seed Foundation Code Team (incl. Shukai Liu)
    arXiv 2024
    [paper]

💻 Experience

  • 2025.09 - Present, IQuest Research, Research Intern, Code LLM Post-training
    • Core contributor to IQuest-Coder-V1 and IQuest-Q1 (open-sourced 320B-A15B MoE, 512K context), responsible for SWE capabilities across SWE-bench Verified / Multilingual / Pro and DeepSWE. IQuest-Q1 reaches 64.6 on DeepSWE v1.1, and IQuest-Coder-V1-40B-Loop reaches 76.2% on SWE-bench Verified.
    • Built the SWE agentic evaluation system with anti-cheating detection, and multi-harness infrastructure (Claude Code, Codex, OpenCode, etc.) shared by evaluation, trajectory distillation and RL / MOPD training.
    • Worked on SWE instance synthesis, high-throughput trajectory distillation, SWE data cleaning, SFT and RL / MOPD training pipelines; synthesized part of the instances for Terminal-Bench 2.1.
  • 2025.01 - 2025.09, Alibaba Group, Future Life Lab of Taobao & Tmall Group, Algorithm Intern
    • Improved code instruction following (+10 pt) and multi-turn code dialogue (+8 pt) of foundation models through data synthesis and benchmark construction; built repository-level issue-resolution data for SWE-bench with the Agentless framework.
  • 2023.11 - 2024.02, Meituan, Algorithm Intern
    • Fine-tuned LLMs with chain-of-thought data for search post-processing in an LLM-based assistant (F1 +~10%), and explored model merging via parameter editing to reduce deployment cost.

📖 Education

  • 2025.09 - Present, Ph.D. Student, School of Computer Science and Engineering, Beihang University. Advisors: Prof. Bo Jiang and Prof. Jian Yang.
  • 2024.09 - 2025.09, Master’s Student (transferred to the Ph.D. program), School of Computer Science and Engineering, Beihang University. Advisor: Prof. Zhoujun Li.
  • 2020.09 - 2024.07, B.E. in Computer Science and Technology, Beijing University of Posts and Telecommunications. GPA 3.83/4.00, ranked 14/442.

🎖 Honors and Awards

  • 2024.06 Outstanding Graduate, Beijing University of Posts and Telecommunications
  • 2023.10 Merit Student, Beijing University of Posts and Telecommunications
  • 2023.04 Third Prize, Lanqiao Cup Programming Contest (C/C++, Group A)
  • 2022.12 Third Prize, Beijing College Students Mathematics Competition
  • 2022.12 Outstanding Communist Youth League Member, BUPT
  • 2022.10 Second-Class Scholarship, BUPT
  • 2021.10 First-Class Scholarship, BUPT