English(EN) Detecting Fluent Optimization-Based Adversarial Prompts via Sequential Entropy Changes

新方法利用序列熵变化检测对抗性LLM提示

作者 PulseAugur 编辑部 · [1 个来源] · 2026-05-19 15:15

研究人员开发了一种名为CPD Online的新方法来检测试图越狱大型语言模型的对抗性提示。该技术将提示检测视为在线变化点检测问题，分析模型令牌预测中的序列熵变化。CPD Online不依赖于特定模型，无需训练，并且能够精确定位恶意提示的开始，在各种开源模型上表现优于现有的困惑度检测器。 AI

影响这种新的检测方法可以通过识别和缓解恶意提示来增强LLM的安全性，从而可能减少对广泛护栏干预的需求。

排序理由该集群包含一篇详细介绍LLM中对抗性提示检测新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

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中对抗性提示检测新方法的学术论文。[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: 141 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。

arXiv cs.AI TIER_1 English(EN) · Miguel R. D. Rodrigues · 2026-05-19 15:15

通过序列熵变化检测基于流畅优化的对抗性提示

Optimization-based adversarial suffixes can jailbreak aligned large language models (LLMs) while remaining fluent, weakening static and windowed perplexity-based detectors. We cast adversarial suffix detection as an online change-point detection problem over the token-level next-…