phenomenon · first seen 4 Oct 2026
jailbreak attack
Also: jailbreaking, LLM jailbreak, prefill attack, refusal bypass, guardrail bypass, LLM jailbreaking, prompt jailbreak
Inputs crafted to bypass a language model's safety training and elicit harmful responses.
Mentions per week
PapersCodeDiscussion
Rank 71 on 16 Nov 2026 · 34 mentions in 28 days · 34 independent groups · papers → code · first seen 6 weeks ago · acceleration -0.37
Earliest sources
- Beyond Refusal Patterns: Safe-Role Internalization for Robust and Generalizable LLM Safety AlignmentPapers · 4 Oct 2026 · Jinghao Pang, Jitai Hao, Qiang Huang, Zhaochun Ren, Jun Yu
- Reward Stealing Attack on Large Language ModelsPapers · 5 Oct 2026 · Jiaming Qian, Pengyang Zhou, Jiahe Xu, Chaochao Chen
- Jailbreaking Open-Weight LLMs via Random Embedding PerturbationsPapers · 5 Oct 2026 · Abhinav Sudhakar Dubey, Scott Sirri, Vaggos Chatziafratis, C. Seshadhri
- Secure Speculative Decoding for Large Language ModelsPapers · 6 Oct 2026 · Yichi Zhang, Zhiqi Wang, Neil Gong, Yuchen Yang
- Quad-State Safety Evaluation of Open-Weight Large Language Models on Non-Canonical InputsPapers · 6 Oct 2026 · Pavan Maddula
- Safe at One Loop, Risky at Another: Aligning Safety Across Recurrent Depths in Looped Language ModelsPapers · 7 Oct 2026 · Yi Wang, Xiuyuan Qi, Dongqi Han, Dongsheng Li, Wenjie Wang
Latest sources
- Hardening Agentic Web Populations Through Multi-Criterion Competitive Co-Evolution: Lessons from Wargame Strategy OptimizationPapers · 14 Nov 2026 · Shashank Raj, Kalyanmoy Deb
- LLMs are Vulnerable to Malicious Requests with Indistinguishable IntentPapers · 13 Nov 2026 · Wenzhuo Xu, Zonghao Ying, Deyue Zhang, Dongdong Yang, Xiangzheng Zhang, Quanchen…
- ramonburf/Tims-CookedCode · 9 Nov 2026 · ramonburf · 53 stars
- 34306/doublera1nCode · 8 Nov 2026 · 34306 · 176 stars
- Enmo Safety: External, Non-invasive, Model-agnostic Oversight for Off-the-Shelf Language ModelsPapers · 6 Nov 2026 · Mohammed Nayeem Khan, Sarun Kumar K. S., Allan Kevin T., Chirag Vaidya, Vrishank…
- META-IR: Meta-Introspective Reasoning for Adaptive Safety Alignment in LLMPapers · 6 Nov 2026 · Mahabubur Rahman Miraj, Hongyu Huang, Nankun Mu, Chao Chen, Shanqin Yi
- Umbraricshadow/gpt4o-jailbreak-2026-respCode · 5 Nov 2026 · Umbraricshadow · 37 stars
- "Blind Compliance": Coercive Control in AI JailbreakingPapers · 5 Nov 2026 · Acacia L. Ackles, Elliott Whitson
- JailBound: FOL-Guided Jailbreak Evaluation for Exploring LLM Behavioral BoundariesPapers · 5 Nov 2026 · Fazong Wu, Xin Wang, Ming Yang, Zhenyong Zhang, Xiaoming Wu
- Addition by Subtraction: Toward Mitigating Fragile Safety Alignment in Large Language ModelsPapers · 4 Nov 2026 · Hamid Dashtbani, Alexandra Dmitrienko
- SafeStart: Mitigating the Overlooked Safety Cost of Active Expert Reduction in MoE LLM ServingPapers · 4 Nov 2026 · Erfan Baghaei Potraghloo, Saeid Shokoufa, Seyedarmin Azizi, Arya Fayyazi, Souvik…
- BrainwashedLLM: Jailbreaking via Automated Adversarial Chain of Thoughts DiscoveryPapers · 3 Nov 2026 · Masoud Poorghaffar Aghdam, Göktuğ Serdar Yıldırım, Sinem Sav