Research Staff
Dr Rui ZHANG
HKU Musketeers Foundation Institute of Data Science
Key Expertise
Trustworthy Machine Learning; Distributed Learning; LLM-based AI Agents
About Me
Dr. Rui Zhang is a Post-doctoral Research Fellow at the HKU Musketeers Foundation Institute of Data Science (HKU IDS), working with Professor Ho Chen. He received his Ph.D. in Computer Science from The Hong Kong Polytechnic University, where he focused on data reconstruction and gradient inversion in Federated Learning. His current research focuses on AI security and privacy, trustworthy machine learning, and LLM-based agents.
Current Research Project
Rui is investigating security vulnerabilities in AI agents, distinguishing between endogenous (model-inherent) and exogenous (external) threats. His work explores multimodal prompt injection attacks that exploit different modalities as carriers, as well as attack-chain hijacking via prompt injection and jailbreaking. He remains open to tackling diverse security and privacy problems in the field of large language models.
Selected Publications
- Junjie Shan, Ziqi Zhao, Jialin Lu, Rui Zhang, Siu Ming Yiu, Ka-Ho Chow. “Geminio: Language-Guided Gradient Inversion Attacks in Federated Learning.” International Conference on Computer Vision (ICCV), 2025.
- Rui Zhang, Song Guo, Ping Li. “GradFilt: Class-wise Targeted Data Reconstruction from Gradients in Federated Learning.” The Web Conference (WWW), 2024.
- Rui Zhang, Song Guo, Ping Li. “Posterior Probability-based Label Recovery Attack in Federated Learning.” ICLR PML Workshop, 2024.
- Rui Zhang, Song Guo, Junxiao Wang, Xin Xie, Dacheng Tao. “A Survey on Gradient Inversion: Attacks, Defenses and Future Directions.” International Joint Conference on Artificial Intelligence (IJCAI), 2022.

