English(EN) Beyond Attack Success Rate: Cost-Aware Metric for Evaluating Iterative LLM Jailbreaks

新指标通过迭代攻击成本评估LLM越狱

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-23 04:00

研究人员开发了一种新指标 ASR_STEP，用于评估迭代式大型语言模型 (LLM) 越狱的有效性。该指标考虑了对受害者模型的调用次数，惩罚需要更多交互的方法。在 JailBreakBench 上对不同模型家族的评估显示，像 GCG 这样的方法在考虑重复调用时得分显著下降，这凸显了考虑迭代攻击成本的重要性。 AI

影响通过考虑迭代攻击的成本，为LLM越狱防御引入了更现实的评估。

排序理由介绍LLM安全新评估指标的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.CL 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 4 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 介绍LLM安全新评估指标的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, safety; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Same-day; Cluster formed today. Ranking reflects the current source set at time of score.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.CL TIER_1 English(EN) · Rohan Menon, Chih-Hong Cheng · 2026-10-23 04:00

Beyond Attack Success Rate: Cost-Aware Metric for Evaluating Iterative LLM Jailbreaks

arXiv:2610.26589v2 Announce Type: replace-cross Abstract: Attack success rate (ASR) evaluates jailbreaks against large language models without accounting for repeated calls to the victim model. Yet each additional call gives the deployed system another opportunity to detect the a…

本文如何被排名

报道来源 [1]

Beyond Attack Success Rate: Cost-Aware Metric for Evaluating Iterative LLM Jailbreaks

相关实体

相关话题