Hu Jinming 胡锦明

Biography

I am a 2022 undergraduate student majoring in Information Security at the School of Computer Science, Shanghai Jiao Tong University, and I expect to graduate in 2026. My current research interests lie in AI security, trustworthy machine learning, federated learning security, and watermarking for language models. I have been recommended for graduate study at SJTU and expect to continue as a master's student in Cyberspace Security in Fall 2026. I am particularly interested in backdoor attacks and defenses, secure federated instruction tuning, and practical security problems in modern foundation-model systems.

News

• [2026.03] EmbTracker: Traceable Black-box Watermarking for Federated Language Models was released on arXiv.
• [2026.03] ProtegoFed: Backdoor-Free Federated Instruction Tuning with Interspersed Poisoned Data was released on arXiv.
• [2026.02] Revisiting Backdoor Threat in Federated Instruction Tuning from a Signal Aggregation Perspective was accepted by ICASSP 2026. 🎉🎉
• [2025] NSmark: Null Space Based Black-box Watermarking Defense Framework for Pre-trained Language Models appeared at the ICLR 1st GenAI Watermarking Workshop. 🎉🎉

Selected Publications

Full publication list is available on Google Scholar.

2026

1. Revisiting Backdoor Threat in Federated Instruction Tuning from a Signal Aggregation Perspective

   Haodong Zhao, Jinming Hu, Gongshen Liu

   ICASSP 2026

2. ProtegoFed: Backdoor-Free Federated Instruction Tuning with Interspersed Poisoned Data

   Haodong Zhao, Jinming Hu, Zhaomin Wu, Zongru Wu, Wei Du, Junyi Hou, Caibei Zhao, Zhuosheng Zhang, Bingsheng He, Gongshen Liu

   arXiv preprint, 2026

3. EmbTracker: Traceable Black-box Watermarking for Federated Language Models

   Haodong Zhao, Jinming Hu, Yijie Bai, Tian Dong, Wei Du, Zhuosheng Zhang, Yanjiao Chen, Haojin Zhu, Gongshen Liu

   arXiv preprint, 2026

2025

1. NSmark: Null Space Based Black-box Watermarking Defense Framework for Pre-trained Language Models

   Haodong Zhao, Jinming Hu, Peixuan Li, Fangqi Li, Jinrui Sha, Tianjie Ju, Peixuan Chen, Zhuosheng Zhang, Gongshen Liu

   ICLR 1st GenAI Watermarking Workshop, 2025

Education

• 2022 - 2026, B.Eng. in Information Security, School of Computer Science, Shanghai Jiao Tong University, Shanghai, China
• Recommended for graduate study at SJTU; expected to begin M.S. in Cyberspace Security in Fall 2026

Honors and Awards

• 2022 - 2023 National Scholarship for Undergraduate Students (本科生国家奖学金 / 国家奖学金)

Contact

If you would like to discuss AI security, trustworthy ML, research opportunities, or possible collaboration, feel free to contact me.

hujinming@sjtu.edu.cn | github.com/jmh2003 | Google Scholar