I am Kedong Xiu, a first-year Ph.D student at Zhejiang University, fortunate to be co-advised by Prof. Tianhang Zheng and Prof. Kui Ren.

My research focuses on the intersection of Large Language Models (LLMs), Multimodal Large Language Models (MLLMs), and Cybersecurity, aiming to build trustworthy and robust AI systems.

Currently, I am interested in the following topics:

• Jailbreaking
• Alignment
• Inversion Attack
• Agent Security

My earlier research focused on Password Security , a topic I may revisit in the future as part of my broader interests. 😎😎😎

News

• [2025/08] 🎉🎉 One paper is accepted by Network and Distributed System Security (NDSS) Symposium 2026!
• [2025/07] 🎉🎉 One paper is accepted by ACM Multimedia (MM) 2025! This work completes when I was a remote intern at New York University advised by Prof. Sai Qian Zhang.

Preprints

(^† denotes equal contribution, ^* denotes the corresponding author.)

• [LLM Security] Dynamic Target Attack

  Authors: Kedong Xiu, Churui Zeng, Tianhang Zheng^*, Xinzhe Huang, Xiaojun Jia, Di Wang, Puning Zhao, Zhan Qin, Kui Ren
  Paper: [arXiv]
  Repo: [GitHub]
  

• [LLM Security] Untargeted Jailbreak Attack

  Authors: Xinzhe Huang, Wenjing Hu, Tianhang Zheng^*, Kedong Xiu, Xiaojun Jia, Di Wang, Zhan Qin, Kui Ren
  Paper: [arXiv]
  Repo: [GitHub]
  

• [LLM Security] HarmMetric Eval: Benchmarking Metrics and Judges for LLM Harmfulness Assessment

  Authors: Langqi Yang, Tianhang Zheng^*, Kedong Xiu, Yixuan Chen, Di Wang, Puning Zhao, Zhan Qin, Kui Ren
  Paper: [arXiv]
  Repo: [GitHub], [Dataset]
  

Publications

(^†Denotes equal contribution.)

Google Scholar

• [LLM Security] DualBreach: Efficient Dual-Jailbreaking via Target-Driven Initialization and Multi-Target Optimization

  Authors: Xinzhe Huang^†, Kedong Xiu^†, Tianhang Zheng^*, Churui Zeng, Wangze Ni, Zhan Qin, Kui Ren, Chun Chen
  Venue: NDSS 2026 (BIG4, CCF-A)
  Paper: [arXiv]
  Repo: [GitHub]
  

• [VLM Security] CapRecover: A Cross-Modality Feature Inversion Attack Framework on Vision Language Models

  Authors: Kedong Xiu and Sai Qian Zhang^*
  Venue: ACM Multimedia (MM) 2025 (CCF-A)
  Paper: [arXiv]
  Repo: [WebPage], [GitHub]

• [Password Security] PointerGuess: Targeted Password Guessing Model Using Pointer Mechanism

  Authors: Kedong Xiu and Ding Wang^*
  Venue: USENIX Security 2024 (BIG4, CCF-A)
  Paper: [USENIX SEC]
  

• [Password Security] Password Guessing using Random Forest

  Authors: Ding Wang^*, Yunkai Zou, Zijian Zhang, Kedong Xiu
  Venue: USENIX Security 2023 (BIG4, CCF-A)
  Paper: [USENIX SEC]