Jingwei Sun
- 310412@stu.xjtu.edu.cn
- Google Scholar
- Github
- Hong Kong, China
I am a Ph.D. student at Trustworthy Machine Learning and Reasoning (TMLR) group in the Department of Computer Science, Hong Kong Baptist University (HKBU), advised by Prof. Bo Han.
Previously, I received both of my MPhil Eng. degree and Bachelor Eng. degree from Xi'an Jiaotong University.
My research interests focus on trustworthy machine learning, particularly in improving the safety, robustness, and reliability of foundation models (LLMs, VLMs), with specific research topics such as adversarial attacks, test-time adaptation, GUI agents, and agent memory. My long-term goal is to build adaptive AI systems that remain dependable as they evolve in complex environments.
I am always open to discussions and possible collaborations. Please feel free to email me if you would like to chat.
News:
- [May 2026] One paper on Robustness Evaluation for GUI Agents has been accepted by ICML 2026.
- [Jan 2026] One paper on Test-Time Adaptation for VLMs has been accepted by ICLR 2026.
- [Mar 2025] One paper on copyright protection against MVDMs has been accepted by ICME 2025.
- [Jul 2024] One paper on physical attack against remote sensing has been accepted by TGRS.
Selected Publication
Conference Paper:
-
AgentHijack: Benchmarking Computer Use Agent Robustness to Common Environment CorruptionsJingwei Sun, Jianing Zhu, Yuanyi Li, Tongliang Liu, Xia Hu, Bo Han#
-
Bilateral Information-aware Test-time Adaptation for Vision-Language ModelsJingwei Sun*, Jianing Zhu*, Jiangchao Yao, Gang Niu, Masashi Sugiyama, Bo Han#
-
Latent Feature and Attention Dual Erasure Attack against Multi-View Diffusion Models for 3D Assets ProtectionJingwei Sun, Xuchong Zhang#, Changfeng Sun, Qicheng Bai, Hongbin Sun
Journal Paper:
-
Physical Strip Attack for Object Detection in Optical Remote SensingChangfeng Sun, Jingwei Sun, Xuchong Zhang#, Yitong Li, Qicheng Bai, Hongbin SunIEEE Transactions on Geoscience and Remote Sensing (TGRS), 2024 [Paper]