利用大语言模型（LLM）攻击受大语言模型保护的文本到图像模型
Harnessing LLM to Attack LLM-Guarded Text-to-Image Models
AI总结：
针对文本到图像模型安全过滤器，提出LLM驱动的多智能体方法DACA，通过将绘图意图重述为视觉组件的良性描述生成对抗性提示，成功绕过DALL-E 3和Midjourney过滤器，单次攻击成功率最高达76.7%和64%，复用攻击达98%和84%，并开源代码数据集。
摘要：
为防止文本到图像（T2I）模型生成不道德图像，人们部署安全过滤器来拦截不当绘图提示。此前研究采用令牌替换搜索对抗性提示以绕过这些过滤器，但因无意义令牌无法通过语义逻辑检查而失效。本文从不同视角处理对抗性提示：将绘图意图重述为单个视觉组件的多个良性描述，可获得有效对抗性提示。我们提出名为DACA的大语言模型驱动多智能体方法，自动完成目标重述。该方法成功绕过DALL-E 3和Midjourney的安全过滤器生成目标图像，单次攻击成功率分别达76.7%和64%，复用攻击成功率分别为98%和84%。代码和数据集已开源。
英文摘要：
To prevent Text-to-Image (T2I) models from generating unethical images, people deploy safety filters to block inappropriate drawing prompts. Previous works have employed token replacement to search adversarial prompts that attempt to bypass these filters, but they have become ineffective as nonsensical tokens fail semantic logic checks. In this paper, we approach adversarial prompts from a different perspective. We demonstrate that rephrasing a drawing intent into multiple benign descriptions of individual visual components can obtain an effective adversarial prompt. We propose a LLM-piloted multi-agent method named DACA to automatically complete intended rephrasing. Our method successfully bypasses the safety filters of DALL-E 3 and Midjourney to generate the intended images, achieving success rates of up to 76.7% and 64% in the one-time attack, and 98% and 84% in the re-use attack, respectively. We open-source our code and dataset on [this link](https://github.com/researchcode003/DACA).
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。
