I am a final-year Ph.D. student in Computer Science and Engineering at The Chinese University of Hong Kong (CUHK), advised by Prof. Michael R. Lyu. My research primarily focuses on Code LLMs and Coding Agents. I work across their training, evaluation, and inference-time optimization, and have published 10+ first-author papers in these areas, including one first-author ICSE 2025 paper that received the ACM SIGSOFT Distinguished Paper Award. Recently, I’m also working on Recursive Self-Improvement, particularly on how to enhance models’ ability to train themselves or optimize their own harness. The overview of my main first-author papers are summarized as below.

🔥 News

• 2026.10: 🎉 Checkout our latest work RSI-Dojo: Training LLM Agents to Automate LLM Post-Training. We construct a suite of post-training environments and teach LLM agents to autonomously construct data, optimize training recipes, and iteratively improve perofrmance. Our RSI-Dojo-35B-A3B reaches 29.95 on PostTrainBench, surpassing the performance of frontier models such as GPT-5.5.
• 2026.08: 🎉 Our paper SWE-MeM: Learning Adaptive Memory Management for Long-Horizon Coding Agents is accepted by EMNLP 2026 as an oral presentation!
• 2025.11: 🎉 Our paper ContextPilot: Code Context Engineering with Memory-Augmented Exploration Agents is accepted by LLM4Code 2026!
• 2025.10: 🎉 Our paper SEER: Enhancing Chain-of-Thought Code Generation through Self-Exploring Deep Reasoning is accepted by ICSE 2026!
• 2025.04: 🏆 Our paper Search-Based LLMs for Code Optimization receives the ACM SIGSOFT Distinguished Paper Award at ICSE 2025!
• 2025.04: 🎉 Beyond PEFT and RAG or Fine-tuning? are accepted by FSE 2025 and the FSE 2025 Industry Track!

📝 Selected Publications

• To appear on arXiv RSI-Dojo: Training LLM Agents to Automate LLM Post-Training, Shuzheng Gao, Zhaojian Yu, Penghao Yin, Shilin He, Kai Cai, and Michael R. Lyu.

• EMNLP 2026 SWE-MeM: Learning Adaptive Memory Management for Long-Horizon Coding Agents, Shuzheng Gao, Wenhao Zeng, Zhaojian Yu, Jianqiao Wangni, Chaozheng Wang, Kai Cai, Shilin He, and Michael R. Lyu. Oral

• ICSE 2026 SEER: Enhancing Chain-of-Thought Code Generation through Self-Exploring Deep Reasoning, Shuzheng Gao, Chaozheng Wang, Cuiyun Gao, and Michael R. Lyu. [Code]

• LLM4Code 2026 ContextPilot: Code Context Engineering with Memory-Augmented Exploration Agents, Shuzheng Gao, Chaozheng Wang, Shuqing Li, Yun Peng, and Michael R. Lyu.

• ICSE 2025 Search-Based LLMs for Code Optimization, Shuzheng Gao, Cuiyun Gao, Wenchao Gu, and Michael R. Lyu. [Code] ACM SIGSOFT Distinguished Paper Award

• ICSE 2024 Learning in the Wild: Towards Leveraging Unlabeled Data for Effectively Tuning Pre-trained Code Models, Shuzheng Gao, Wenxin Mao, Cuiyun Gao, Li Li, Xing Hu, Xin Xia, and Michael R. Lyu. [Code]

• ASE 2023 What Makes Good In-context Demonstrations for Code Intelligence Tasks with LLMs?, Shuzheng Gao, Xin-Cheng Wen, Cuiyun Gao, Wenxuan Wang, Hongyu Zhang, and Michael R. Lyu. [Code]

• ICSE 2023 Keeping Pace with Ever-Increasing Data: Towards Continual Learning of Code Intelligence Models, Shuzheng Gao, Hongyu Zhang, Cuiyun Gao and Chaozheng Wang. [Code]

• ICSE 2023 Two Sides of the Same Coin: Exploiting the Impact of Identifiers in Neural Code Comprehension, Shuzheng Gao, Cuiyun Gao, Chaozheng Wang, Jun Sun, David Lo, and Yue Yu. [Code]

• TOSEM 2022 Code Structure Guided Transformer for Source Code Summarization, Shuzheng Gao, Cuiyun Gao, Yulan He, Jichuan Zeng, Lun Yiu Nie, Xin Xia, and Michael R. Lyu. [Code]

^* indicates equal contribution. See Publications for the complete list.

💻 Internships

• Dec 2025 - Present, Research Intern, ByteDance Douyin AI4SE.
  • Mentor: Dr. Shilin He.
  • Research Focus: Coding Agent SFT and RL. Especially focusing on: (1) Design and train long-horizon coding agents with compaction and subagents. (2) Explore agentic RL training recipe and credit assignment methods. (3) Research on automated post-training.
• Mar 2021 - Aug 2021, Research Intern, Huawei Noah’s Ark Lab.
  • Mentor: Dr. Pengyun Wang.
  • Research Focus: Long-tailed image classification.

🎖 Honors and Awards

• 2025: ACM SIGSOFT Distinguished Paper Award, ICSE.
• 2025: Distinguished Reviewer Award, Internetware.
• 2023: Distinguished Paper Award, ASE Industry Challenge.
• 2023: Binxing Fang Scholarship.
• 2022: Best Student Paper Award, ACAIT.
• 2021: Outstanding Graduate of Harbin Institute of Technology.
• 2019: National Scholarship, Ministry of Education.