Defending Jailbreak Attacks on Large Language Models via Manifold Trajectory Kinetics

Computer Science > Cryptography and Security [Submitted on 5 Jun 2026] Defending Jailbreak Attacks on Large Language Models via Manifold Trajectory Kinetics Hangtao Zhang, Yucheng Zhao, Sishun Liu, Ziqi Zhou, Zeyu Ye, Wei Wan, Minghui Li, Shengshan Hu, Yanjun Zhang, Yi Liu, Leo Yu Zhang Jailbreak prompts can bypass alignment guardrails in large language models (LLMs) and elicit unsafe outputs, making reliable deployment-time detection critical. Prior detection approaches largely rely on a fixed metric space, e.g., raw inputs, gradients, or hidden features, in which benign and jailbreak prompts are linearly separable. We show this assumption breaks under (i) pseudo-malicious prompts that are benign by intent but contain safety-related keywords, and (ii) adaptive attacks that explicitly optimize against the deployed detector. To overcome this limitation, we shift our focus from identifying a universal metric space to analyzing the more robust neighborhood structure of the underlying data manifold. We present Manifold Trajectory Kinetics (MTK), which treats an LLM as a kinetic system transforming inputs into outputs and detects jailbreaks by tracking how a prompt's neighborhood structure evolves across layers. Benign prompts remain close to benign neighborhoods throughout inference, whereas jailbreak prompts exhibit a characteristic trajectory that begins near malicious seeds and later strategically shifts toward benign neighborhoods to evade this http URL four LLMs and ten jailbreak attacks, MTK achieves strong robustness to both failure modes: on pseudo-malicious prompts, it attains a jailbreak true positive rate of 95% at a false positive rate of 5% on benign prompts and 2% on pseudo-malicious prompts, and under adaptive attacks, it maintains a true positive rate of 85%. We further demonstrate the superior performance of MTK for jailbreak detection in vision-language models. Our code is available at this https URL. Comments: Accepted to USENIX Security '26 Cycle 2. Code is available at this https URL Subjects: Cryptography and Security (cs.CR) Cite as: arXiv:2606.07335 [cs.CR]   (or arXiv:2606.07335v1 [cs.CR] for this version)   https://doi.org/10.48550/arXiv.2606.07335 Focus to learn more Submission history From: Hangtao Zhang [view email] [v1] Fri, 5 Jun 2026 14:49:26 UTC (14,557 KB) Access Paper: HTML (experimental) view license Current browse context: cs.CR < prev   |   next > new | recent | 2026-06 Change to browse by: cs References & Citations NASA ADS Google Scholar Semantic Scholar Export BibTeX Citation Bookmark Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Demos Related Papers About arXivLabs Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)