见多识广？评估面向Agent-to-Agent多模态说服的视觉语言模型易受性
Seeing is Believing? Evaluating Vision-Language Model Susceptibility in Agent-to-Agent Multimodal Persuasion
AI总结：
本文研究了在多智能体多模态说服场景中，视觉语言模型对多模态内容的易受性，提出了MMPersuade框架和数据集，通过实验揭示了多模态输入在说服中的优势，以及说服对象的领域和格式依赖性，以及心理策略在不同上下文和模型架构下的效果差异。
摘要：
随着自主代理越来越多地互动，它们不可避免地试图互相影响。尽管先前在纯文本环境下研究了Agent-to-Agent (A2A) 说服的动力学，但视觉语言模型 (VLMs) 的兴起带来了更复杂的挑战：多模态内容传达了更丰富的信息，同时整合了微妙且难以检测的说服线索。为了研究这种易受性，我们提出了MMPersuade，一个统一的框架和数据集用于A2A多模态说服。我们建模了说服者代理（利用图像和心理策略）与说服对象VLM之间的互动。我们的基准涵盖商业、主观和行为，以及对抗性情境，并通过功能调用评估说服，以捕捉超出口头回应的行为变化。在六个VLM上的实验揭示了三个发现：（1）多模态输入在说服中始终优于纯文本说服，原始视觉信号在对抗性情境中独特地增加易受性，通过绕过文本激活的安全防御；（2）说服对象的易受性高度依赖于领域和格式，现实和社区风格的格式在商业情境中驱动易受性，而不同格式在对抗性情境中占主导地位；（3）心理策略的有效性取决于上下文和模型架构，更强大的模型抵抗良性说服，但在对抗性多模态输入下更易受攻击。我们的框架为构建更稳健和对齐的VLMs提供了基础，以在多代理环境中使用。
英文摘要：
As autonomous agents increasingly interact, they inevitably attempt to influence one another. While prior work in text-only settings has explored the dynamics of Agent-to-Agent (A2A) persuasion, the rise of Vision-Language Models (VLMs) introduces a more complex challenge: multimodal content conveys richer information while integrating subtle, hard-to-detect persuasive cues. To study this vulnerability, we present MMPersuade, a unified framework and dataset for A2A multimodal persuasion. We model interactions between a persuader agent, which leverages images and psychological strategies, and a persuadee VLM. Our benchmark spans commercial, subjective and behavioral, and adversarial contexts, and evaluates persuasion via function-calling that capture behavioral shifts beyond verbal responses. Experiments on six VLMs reveal three findings: (1) multimodal inputs consistently outperform text-only persuasion, with raw visual signals uniquely increasing susceptibility in adversarial settings by bypassing text-activated safety defenses; (2) persuadee vulnerability is highly domain- and format-dependent, with realistic and community-style formats driving susceptibility in commercial settings while different formats dominate in adversarial ones; and (3) psychological strategy efficacy varies with context and model architecture, as more capable models resist benign persuasion yet become more susceptible under adversarial multimodal inputs. Our framework provides a foundation for building more robust and aligned VLMs in multi-agent environments.
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。
