Fudan University
Ph.D. in Computer Science and Technology.
Hi! I am an incoming Ph.D. student in Computer Science and Technology at Fudan University, starting in September 2026, advised by Chaochao Lu with joint advisors Xiang Wang and YouBang Sun, and previously by Jie Fu. I received my M.S. from Shanghai Jiao Tong University, where I was supervised by Kai Wang as a member of Xuemin Lin's group. I received my B.Eng. from East China Normal University, advised by Lyu Ni and Xiang Li. My current research focuses on trustworthy AI systems, particularly self-evolving AI agents and their supporting architectures and infrastructure.
Education
2026 -
2023 - 2026
Shanghai Jiao Tong University
M.S. in Management Science and Engineering.
Thesis: An Interactive Community Search Framework via Graph Active Learning.
2019 - 2023
East China Normal University
B.Eng. in Data Science and Big Data Technology.
Thesis: Cryptocurrency Portfolio Management Based on Deep Reinforcement Learning.
News
- 08/2026Modular Representation Steering and EvoSQL were accepted to EMNLP 2026 Main and Findings, respectively.
- 07/2026Presented ARK as an oral paper at ACL 2026.
- 06/2026Selected as an Outstanding Graduate of Shanghai Jiao Tong University.
- 05/2026TrustResearcher presentation changed to virtual due to conflicts. Stay safe!
- 05/2026ARK selected as an ACL 2026 oral presentation.
- 04/2026ARK accepted to ACL 2026.
- 01/2026TrustResearcher accepted as a WWW 2026 Demo. See you in Dubai!
- 10/2025Received the National Scholarship.
- 08/2025Started as a research intern at Shanghai AI Laboratory.
- 06/2025COMET accepted by INFORMS Journal on Computing.
- 03/2025Admitted to the joint Ph.D. program between Shanghai AI Laboratory and Fudan University.
Research
My research focuses on trustworthy AI systems, including self-evolving AI agents, efficient agentic RAG, activation steering, and post-training for long-horizon code agents. I aim to develop continuous-control harnesses that coordinate self-evolution across multiple timescales, from fast agent-level adaptation to slower model-level updates, enabling continual learning from long-horizon interaction and feedback. Any form of constructive feedback and collaboration is welcome; feel free to send me an email.
Selected Publications
* Equal contribution. † Co-corresponding authors.
Academic Service
Reviewer
2026
ACL Rolling Review (ARR); AAAI
2025
AAAI
Teaching Assistant
2024
Programming (BUSS1301), Shanghai Jiao Tong University
Data Structures (MGTS3401), Shanghai Jiao Tong University
Internship
08/2025 - Present
Research Intern, Shanghai AI Laboratory
Working on self-evolving coding agents and co-evolving architectures spanning models, harnesses, and activation-steering mechanisms. I study the mechanisms behind post-training recipes, with the goal of unifying self-evolution, self-distillation, and self-play to build agentic systems capable of recursive self-improvement (RSI).
04/2025 - 10/2025
Algorithm Engineer, WANO Shanghai
Built WANO-Agent, a collaborative multi-agent framework based on MCP and DeepSeek-V3 for internal peer-review workflows, with a Milvus-backed knowledge base.
07/2022 - 09/2022
NLP Intern, CloudWalk
Implemented a BERT-BiLSTM-CRF prosodic structure prediction pipeline and introduced self-training plus active learning on internal data.
Honors
Master
National Scholarship (2025); First-Class Scholarship.
Undergraduate
National Scholarship (2022); First-Class Scholarship; Meritorious Winner in MCM/ICM.