面向大语言模型的性格编辑
Editing Personality for Large Language Models
AI总结：
本文提出大语言模型性格编辑任务，构建 PersonalityEdit 基准数据集，基于社会心理学选取神经质、外向性和宜人性三种特质，通过 GPT-4 生成数据并进行多基线实验，揭示该任务面临的挑战。
摘要：
本文提出了一项创新任务，专注于编辑大语言模型（LLMs）的性格特质。该任务旨在调整模型在特定主题上对观点相关问题的回答，因为个体的性格通常以其表达的观点形式体现，从而展现出不同的性格特质。具体而言，我们构建了 PersonalityEdit，一个用于解决该任务的新基准数据集。借鉴社会心理学理论，我们分离出三种代表性特质，即神经质、外向性和宜人性，作为该基准的基础。随后，我们使用 GPT-4 收集数据，生成与指定主题一致并体现目标性格特质的回答。我们进行了涉及多种基线的综合实验，并讨论了大语言模型中性格行为的表征。我们的发现揭示了所提任务的潜在挑战，说明了若干尚待解决的问题。我们期望我们的工作能够推动模型编辑和性格相关研究中的进一步标注工作。代码可在 https://github.com/zjunlp/EasyEdit 获取。
英文摘要：
This paper introduces an innovative task focused on editing the personality traits of Large Language Models (LLMs). This task seeks to adjust the models' responses to opinion-related questions on specified topics since an individual's personality often manifests in the form of their expressed opinions, thereby showcasing different personality traits. Specifically, we construct PersonalityEdit, a new benchmark dataset to address this task. Drawing on the theory in Social Psychology, we isolate three representative traits, namely Neuroticism, Extraversion, and Agreeableness, as the foundation for our benchmark. We then gather data using GPT-4, generating responses that align with a specified topic and embody the targeted personality trait. We conduct comprehensive experiments involving various baselines and discuss the representation of personality behavior in LLMs. Our findings uncover potential challenges of the proposed task, illustrating several remaining issues. We anticipate that our work can stimulate further annotation in model editing and personality-related research. Code is available at https://github.com/zjunlp/EasyEdit.
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。
