English(EN) D-Judge: Disrupting Multi-Turn Jailbreaks using Semantics-Preserving Output Rewriting

新的D-Judge防御通过输出重写来破坏LLM越狱

作者 PulseAugur 编辑部 · [1 个来源] · 2026-06-03 04:00

研究人员开发了一种名为D-Judge的新防御机制，以对抗大型语言模型的多轮越狱攻击。这些攻击利用辅助裁判模型提供的反馈，通过迭代优化提示以实现有害目标。D-Judge通过在受害者LLM的响应被攻击者的裁判评估之前对其进行重写来工作，从而在不改变响应含义的情况下使反馈信号失配。这种策略会扰乱提示优化过程，从而在HarmBench等基准测试中提高安全性，同时保持在良性任务上的性能。 AI

影响引入了一种针对复杂多轮越狱的新型防御措施，有望提高LLM的安全性和可靠性。

排序理由该集群包含一篇详细介绍LLM安全新型防御机制的研究论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 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: 0 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群包含一篇详细介绍LLM安全新型防御机制的研究论文。[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: safety, paper; 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: 128 days old; Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Huanli Gong, Zhipeng Wei, Yu Fu, Haz Sameen Shahgir, Ananya Gupta, Yue Dong, N. Benjamin Erichson · 2026-06-03 04:00

D-Judge：利用语义保留输出重写来扰乱多轮越狱

arXiv:2606.02640v1 Announce Type: cross Abstract: Multi-turn jailbreak attacks pose a growing threat to large language model (LLM) safety because they exploit feedback from auxiliary judge models to iteratively refine prompts toward harmful goals. Existing defenses largely detect…

本文如何被排名

报道来源 [1]

D-Judge：利用语义保留输出重写来扰乱多轮越狱

相关实体

相关话题