cs.AI updates on arXiv.org 10月07日
AgentTypo:对抗视觉语言模型的新型黑盒攻击框架
index_new5.html
../../../zaker_core/zaker_tpl_static/wap/tpl_guoji1.html

 

本文提出了一种名为AgentTypo的黑盒攻击框架,通过将优化文本嵌入网页图像,实现了自适应的排版式提示注入。实验结果表明,AgentTypo在多种场景下均优于现有的图像攻击方法,对多模态智能体构成实际威胁。

arXiv:2510.04257v1 Announce Type: cross Abstract: Multimodal agents built on large vision-language models (LVLMs) are increasingly deployed in open-world settings but remain highly vulnerable to prompt injection, especially through visual inputs. We introduce AgentTypo, a black-box red-teaming framework that mounts adaptive typographic prompt injection by embedding optimized text into webpage images. Our automatic typographic prompt injection (ATPI) algorithm maximizes prompt reconstruction by substituting captioners while minimizing human detectability via a stealth loss, with a Tree-structured Parzen Estimator guiding black-box optimization over text placement, size, and color. To further enhance attack strength, we develop AgentTypo-pro, a multi-LLM system that iteratively refines injection prompts using evaluation feedback and retrieves successful past examples for continual learning. Effective prompts are abstracted into generalizable strategies and stored in a strategy repository, enabling progressive knowledge accumulation and reuse in future attacks. Experiments on the VWA-Adv benchmark across Classifieds, Shopping, and Reddit scenarios show that AgentTypo significantly outperforms the latest image-based attacks such as AgentAttack. On GPT-4o agents, our image-only attack raises the success rate from 0.23 to 0.45, with consistent results across GPT-4V, GPT-4o-mini, Gemini 1.5 Pro, and Claude 3 Opus. In image+text settings, AgentTypo achieves 0.68 ASR, also outperforming the latest baselines. Our findings reveal that AgentTypo poses a practical and potent threat to multimodal agents and highlight the urgent need for effective defense.

Fish AI Reader

Fish AI Reader

AI辅助创作,多种专业模板,深度分析,高质量内容生成。从观点提取到深度思考,FishAI为您提供全方位的创作支持。新版本引入自定义参数,让您的创作更加个性化和精准。

FishAI

FishAI

鱼阅,AI 时代的下一个智能信息助手,助你摆脱信息焦虑

联系邮箱 441953276@qq.com

相关标签

多模态智能体 黑盒攻击 视觉语言模型 AgentTypo 排版式提示注入
相关文章