警惕环境影响：多模态大语言模型智能体易受环境干扰
Caution for the Environment: Multimodal LLM Agents are Susceptible to Environmental Distractions
AI总结：
本文研究GUI环境中多模态大语言模型智能体的忠实性，发现通用和专用GUI智能体均易受环境无关内容干扰，还实现了对抗性环境注入并分析提升方法，呼吁关注该问题。
摘要：
本文研究了图形用户界面（GUI）环境中多模态大语言模型（MLLM）智能体的忠实性，旨在解决多模态GUI智能体是否会被环境上下文干扰的研究问题。本文提出了一种通用场景：用户与智能体均为良性，环境虽无恶意但包含无关内容。研究采用模拟数据集，按照三种感知程度不同的工作模式，评估了多种MLLM作为GUI智能体的表现。实验结果显示，即便是性能最强的模型，无论是通用智能体还是专用GUI智能体，都易受干扰。当前研究大多聚焦于智能体的有用性，而本文发现首次表明这类智能体易受环境干扰。此外，本文实现了对抗性环境注入，并分析了提升忠实性的方法，呼吁学界共同关注这一重要课题。
英文摘要：
This paper investigates the faithfulness of multimodal large language model (MLLM) agents in a graphical user interface (GUI) environment, aiming to address the research question of whether multimodal GUI agents can be distracted by environmental context. A general scenario is proposed where both the user and the agent are benign, and the environment, while not malicious, contains unrelated content. A wide range of MLLMs are evaluated as GUI agents using a simulated dataset, following three working patterns with different levels of perception. Experimental results reveal that even the most powerful models, whether generalist agents or specialist GUI agents, are susceptible to distractions. While recent studies predominantly focus on the helpfulness of agents, our findings first indicate that these agents are prone to environmental distractions. Furthermore, we implement an adversarial environment injection and analyze the approach to improve faithfulness, calling for a collective focus on this important topic.
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。
