He Du / 杜贺
I am a third-year Ph.D. student at the School of Computer Science and Technology, Fudan University, advised by Prof. Dacheng Tao. I completed my undergraduate studies at Harbin Institute of Technology.
My research focuses on agent coding, RL, and data synthesis. The three are hard to separate: a coding agent is only as good as the feedback it can act on, RL needs that feedback to be dense and checkable, and synthesis is how you get enough of it without leaning on human labels. I am interested in closing this loop so models can improve at writing and repairing real software, rather than only solving static benchmarks.
Feel free to reach out at elyndendu@gmail.com. You can also download my CV (PDF).
News
- 2026.08Qwen3.8-Max is released; I contributed to its delivery at Qwen.
- 2026.07EvoSyn is accepted to ACL 2026.
- 2026.06TESSY is accepted to ICML 2026.
- 2026.06Kernel-Smith is accepted to the DL4C workshop at ICML 2026 — state-of-the-art on KernelBench with the Triton backend, plus upstream contributions to SGLang and LMDeploy.
- 2026.05Presented CRew at ICASSP 2026 in Barcelona.
- 2025.10EvoSyn is out; code released at kinza99/openevolve.
- 2025.05SWE-Fixer is accepted to ACL 2025 Findings.
- 2025.01ProverQA is accepted to ICLR 2025.
Publications
2026
2025
Experience
-
Qwen, Alibaba 2026.04 – PresentResearch Intern
-
Shanghai AI Laboratory 2024.09 – 2026.04Research Intern, InternLM Team
Education
- Fudan University 2024.09 – Present
Ph.D. in Artificial Intelligence, School of Computer Science and Technology
- Harbin Institute of Technology 2020.09 – 2024.06
B.Eng. in Computer Science and Technology
Honors & Awards
- 2024Academic Scholarship, Fudan University
- 2024Outstanding Graduate, Harbin Institute of Technology