大型语言模型如何说服？线性探针可揭示多轮对话中的说服动态
How Do LLMs Persuade? Linear Probes Can Uncover Persuasion Dynamics in Multi-Turn Conversations
AI总结：
本研究利用线性探针分析LLMs在多轮对话中的说服动态，证明其能高效识别说服成功点、个性与策略，且优于提示方法，为研究复杂行为提供新途径。
摘要：
大型语言模型（LLMs）已开始展现出说服人类的能力，然而我们对这种动态如何发生理解有限。近期工作使用线性探针（一种用于分析模型表示的轻量级工具）来研究LLMs的各种技能，如建模用户情绪和政治观点的能力。受此启发，我们应用探针来研究自然多轮对话中的说服动态。我们利用认知科学的见解，针对说服的不同方面训练探针：说服成功、被说服者个性和说服策略。尽管这些探针简单，但我们表明它们能在样本和数据集层面捕捉说服的多个方面。例如，探针能识别对话中被说服者被说服的时间点，或在整个数据集中说服成功通常发生的位置。我们还表明，除了比昂贵的基于提示的方法更快外，探针在某些情境下（如揭示说服策略时）表现同样出色，甚至优于提示方法。这表明探针是研究其他复杂行为（如欺骗和操纵）的可行途径，尤其是在多轮设置和大规模数据集分析中，基于提示的方法在计算上会低效。
英文摘要：
Large Language Models (LLMs) have started to demonstrate the ability to persuade humans, yet our understanding of how this dynamic transpires is limited. Recent work has used linear probes, lightweight tools for analyzing model representations, to study various LLM skills such as the ability to model user sentiment and political perspective. Motivated by this, we apply probes to study persuasion dynamics in natural, multi-turn conversations. We leverage insights from cognitive science to train probes on distinct aspects of persuasion: persuasion success, persuadee personality, and persuasion strategy. Despite their simplicity, we show that they capture various aspects of persuasion at both the sample and dataset levels. For instance, probes can identify the point in a conversation where the persuadee was persuaded or where persuasive success generally occurs across the entire dataset. We also show that in addition to being faster than expensive prompting-based approaches, probes can do just as well and even outperform prompting in some settings, such as when uncovering persuasion strategy. This suggests probes as a plausible avenue for studying other complex behaviours such as deception and manipulation, especially in multi-turn settings and large-scale dataset analysis where prompting-based methods would be computationally inefficient.
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。
