← 研踪 PaperTrailarXiv:2611.17818 · v1
LLMs are Vulnerable to Malicious Requests with Indistinguishable Intent
首次提交 2026-11-13 14:56
版本更新 2026-11-13 14:56
雷达观测 2026-11-16 10:40
摘要
Jailbreak attacks can circumvent the safety mechanisms of Large Language Models (LLMs) and lead to the generation of harmful or policy-violating content. We first examine eight jailbreak methods against an LLM-based filter and observe that requests resembling normal user queries tend to have higher filter pass rates than those containing obvious anomalous patterns. This observation motivates the hypothesis that resemblance to normal requests may contribute to jailbreak success. We introduce \textbf{Detail Guidance Attack (DGA)}, a framework that uses an auxiliary LLM to construct detailed requests and interacts with the victim model in a single turn. Across 11 victim models and three datasets, the complete DGA pipeline achieves high attack success rates under our automated judging criterion, including rates above 95\% in several evaluated settings. We also evaluate a DGA-aware filter and conduct a user study on everyday requests with ambiguous or dual-use intent. The survey documents differing expectations about whether models should respond; the usability costs of the filter remain to be measured.
观测到的版本
相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。
v1 · 2026-11-13 14:56 · 最新观测版本
LLMs are Vulnerable to Malicious Requests with Indistinguishable Intent
Jailbreak attacks can circumvent the safety mechanisms of Large Language Models (LLMs) and lead to the generation of harmful or policy-violating content. We first examine eight jailbreak methods against an LLM-based filter and observe that requests resembling normal user queries tend to have higher filter pass rates than those containing obvious anomalous patterns. This observation motivates the hypothesis that resemblance to normal requests may contribute to jailbreak success. We introduce \textbf{Detail Guidance Attack (DGA)}, a framework that uses an auxiliary LLM to construct detailed requests and interacts with the victim model in a single turn. Across 11 victim models and three datasets, the complete DGA pipeline achieves high attack success rates under our automated judging criterion, including rates above 95\% in several evaluated settings. We also evaluate a DGA-aware filter and conduct a user study on everyday requests with ambiguous or dual-use intent. The survey documents differing expectations about whether models should respond; the usability costs of the filter remain to be measured.
查看此版本解读 →查看此版本原文 ↗