arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2502.14744cs.CL
收藏

HiddenDetect:通过监控隐藏状态检测针对大型视觉-语言模型的越狱攻击

HiddenDetect: Detecting Jailbreak Attacks against Large Vision-Language Models via Monitoring Hidden States

MMLab, The Chinese University of Hong Kong(中文大学香港大学);Future Lab, Alibaba Group(阿里集团未来实验室)

Yilei Jiang, Xinyan Gao, Tianshuo Peng, Yingshui Tan, Xiaoyong Zhu, Bo Zheng, Xiangyu Yue

AI总结:

HiddenDetect通过监控大型视觉-语言模型的隐藏状态,利用其内在的安全相关激活模式,无需微调即可高效检测越狱攻击,超越现有方法,增强模型鲁棒性。

摘要:

与仅处理语言的模型相比,集成额外模态增加了大型视觉-语言模型(LVLMs)对安全风险(如越狱攻击)的敏感性。尽管现有研究主要关注事后对齐技术,但LVLMs内部潜在的安全机制在很大程度上仍未得到探索。在这项工作中,我们研究了LVLMs在推理过程中是否在其内部激活中固有地编码了与安全相关的信号。我们的发现表明,LVLMs在处理不安全提示时表现出不同的激活模式,这可用于检测和缓解对抗性输入,而无需进行大量微调。基于这一见解,我们引入了HiddenDetect,一种新颖的无调优框架,利用内部模型激活来增强安全性。实验结果表明,HiddenDetect在检测针对LVLMs的越狱攻击方面超越了最先进的方法。通过利用内在的安全感知模式,我们的方法提供了一种高效且可扩展的解决方案,以增强LVLMs对多模态威胁的鲁棒性。我们的代码将在https://github.com/leigest519/HiddenDetect公开发布。

英文摘要:

The integration of additional modalities increases the susceptibility of large vision-language models (LVLMs) to safety risks, such as jailbreak attacks, compared to their language-only counterparts. While existing research primarily focuses on post-hoc alignment techniques, the underlying safety mechanisms within LVLMs remain largely unexplored. In this work , we investigate whether LVLMs inherently encode safety-relevant signals within their internal activations during inference. Our findings reveal that LVLMs exhibit distinct activation patterns when processing unsafe prompts, which can be leveraged to detect and mitigate adversarial inputs without requiring extensive fine-tuning. Building on this insight, we introduce HiddenDetect, a novel tuning-free framework that harnesses internal model activations to enhance safety. Experimental results show that {HiddenDetect} surpasses state-of-the-art methods in detecting jailbreak attacks against LVLMs. By utilizing intrinsic safety-aware patterns, our method provides an efficient and scalable solution for strengthening LVLM robustness against multimodal threats. Our code will be released publicly at https://github.com/leigest519/HiddenDetect.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑