ACL Findings2025
A Mousetrap: Fooling Large Reasoning Models for Jailbreak with Chain of Iterative Chaos
Yang Yao, Xuan Tong, Ruofan Wang, Yixu Wang, Lujundong Li, Liang Liu, Yan Teng, Yingchun Wang
Abstract
Large Reasoning Models (LRMs) have significantly advanced beyond traditional Large Language Models (LLMs) with their exceptional logical reasoning capabilities, yet these improvements introduce heightened safety risks. When subjected to jailbreak attacks, their ability to generate more targeted and organized content can lead to greater harm. Although some studies claim that reasoning enables safer LRMs against existing LLM attacks, they overlook the inherent flaws within the reasoning process itself. To address this gap, we propose the first jailbreak attack targeting LRMs, exploiting their unique vulnerabilities stemming from the advanced reasoning capabilities. Specifically, we introduce a CHAOS MACHINE, a novel component to transform attack prompts with diverse one-to-one mappings. The chaos mappings iteratively generated by the machine are embedded into the reasoning chain, which strengthens the variability and complexity and also promotes a more robust attack. Based on this, we construct the MOUSETRAP framework, which makes attacks projected into nonlinearlike low sample spaces with mismatched generalization enhanced. Also, due to the more competing objectives, LRMs gradually maintain the inertia of unpredictable iterative reasoning and fall into our trap. Success rates of the Mousetrap attacking o1-mini, Claude-Sonnet and Gemini-Thinking are as high as 96%, 86% and 98% respectively on our toxic dataset TROTTER. On benchmarks such as AdvBench, StrongREJECT, and HarmBench, attacking Claude-Sonnet, well-known for its safety, Mousetrap can astonishingly achieve success rates of 87.5%, 86.58% and 93.13% respectively. Attention: This paper contains inappropriate, offensive and harmful content.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
19 top-tier papers have cited it since. Know what they found before your reviewers do.
Lune can draw on
the full paper · the top-tier papers that cite it · related work
Free to start. No credit card required.
Lineage
- Self-Consistency Improves Chain of Thought Reasoning in Language Models cites Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
- Chain-of-Thought Prompting Elicits Reasoning in Large Language Models cites Self-Consistency Improves Chain of Thought Reasoning in Language Models
- Reasoning as an Adaptive Defense for Safety cites HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
- Reasoning as an Adaptive Defense for Safety cites Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
- Reasoning as an Adaptive Defense for Safety cites AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language Models
- AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language Models cites HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Cited by top-tier papers19
- SafeDialBench: A Fine-Grained Safety Evaluation Benchmark for Large Language Models in Multi-Turn Dialogues with Diverse Jailbreak AttacksHongye Cao, Sijia Jing, Yanming Wang, Ziyue Peng et al.ICLR 2026 · 28 citations
- Reasoning as an Adaptive Defense for SafetyTaeyoun Kim, Fahim Tajwar, Aditi Raghunathan, Aviral KumarNeurIPS 2025 · 24 citations
- Self-Jailbreaking: Language Models Can Reason Themselves Out of Safety Alignment After Benign Reasoning TrainingZheng Xin Yong, Stephen H. BachICLR 2026 · 10 citations
- From Completion to Editing: Unlocking Context-Aware Code Infilling via Search-and-Replace Instruction TuningJiajun Zhang, Zeyu Cui, Jiaxi Yang, Lei Zhang et al.ACL 2026 · 8 citations
- TRAP: Hijacking VLA CoT-Reasoning via Adversarial PatchesZhengxian Huang, Wenjun Zhu, Haoxuan Qiu, Xiaoyu Ji et al.ICML 2026 · 5 citations
Builds on16
- Chain-of-Thought Prompting Elicits Reasoning in Large Language ModelsJason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma et al.NeurIPS 2022 · 22,562 citations
- HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust RefusalMantas Mazeika, Long Phan, Xuwang Yin, Andy Zou et al.ICML 2024 · 1,031 citations
- AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language ModelsXiaogeng Liu, Nan Xu, Muhao Chen, Chaowei XiaoICLR 2024 · 722 citations
- Self-Consistency Improves Chain of Thought Reasoning in Language ModelsXuezhi Wang, Jason Wei, Dale Schuurmans, Quoc V. Le et al.ICLR 2023 · 681 citations
- Catastrophic Jailbreak of Open-source LLMs via Exploiting GenerationYangsibo Huang, Samyak Gupta, Mengzhou Xia, Kai Li et al.ICLR 2024 · 481 citations
Related papers
- HAUNTATTACK: When Attack Follows Reasoning as a ShadowJingyuan Ma, Rui Li, Zheng Li, Junfeng Liu et al.ACL Findings 2026
- Three Minds, One Legend: Jailbreak Large Reasoning Model with Adaptive Stacked CiphersViet-Anh Nguyen, Shiqian Zhao, Gia Dao, Runyi Hu et al.ACL Findings 2026
- VisCRA: A Visual Chain Reasoning Attack for Jailbreaking Multimodal Large Language ModelsBingrui Sima, Linhua Cong, Wenxuan Wang, Kun HeEMNLP 2025 · 11 citations
- Reasoning-Augmented Conversation for Multi-Turn Jailbreak Attacks on Large Language ModelsZonghao Ying, Deyue Zhang, Zonglei Jing, Yisong Xiao et al.EMNLP Findings 2025 · 6 citations
- VRSA: Jailbreaking Multimodal Large Language Models through Visual Reasoning Sequential AttackShiji Zhao, Shukun Xiong, Yao Huang, Jin Yan, Zhenyu Wu, Jiyang Guan, Ranjie Duan, Jialing Tao, Hui Xue, Xingxing WeiCVPR Findings 2026
