大五人格特征在大语言模型中的探测与引导：一项大五研究
Linear Personality Probing and Steering in LLMs: A Big Five Study
AI总结：
本文研究了通过大五人格特质对齐的线性方向探测和引导大语言模型行为的有效性，发现其在特定任务中表现良好，但在开放式生成中效果有限。
摘要：
大型语言模型（LLMs）表现出独特且一致的人格特征，这些特征对信任和参与度有重大影响。尽管这表明人格框架将是高度有价值的工具，用于表征和控制LLM的行为，但当前方法要么成本高（训练后），要么脆弱（提示工程）。通过线性方向进行探测和引导最近出现了一种廉价且高效的替代方法。在本文中，我们研究了与大五人格特质对齐的线性方向是否可用于探测和引导模型行为。使用Llama 3.3 70B，我们生成了406个虚构角色的描述及其大五特质分数。然后，我们用这些描述和来自Alpaca问卷的问题提示模型，使我们能够采样沿人格特质变化的隐藏激活。使用线性回归，我们学习了一组每层激活空间中的方向，并测试它们对探测和引导模型行为的有效性。我们的结果表明，与特质分数对齐的线性方向是有效的人格检测探针，而其引导能力则强烈依赖于上下文，在选择性任务中产生可靠的效果，但在开放式生成或提示中存在额外上下文时影响有限。
英文摘要：
Large language models (LLMs) exhibit distinct and consistent personalities that greatly impact trust and engagement. While this means that personality frameworks would be highly valuable tools to characterize and control LLMs' behavior, current approaches remain either costly (post-training) or brittle (prompt engineering). Probing and steering via linear directions has recently emerged as a cheap and efficient alternative. In this paper, we investigate whether linear directions aligned with the Big Five personality traits can be used for probing and steering model behavior. Using Llama 3.3 70B, we generate descriptions of 406 fictional characters and their Big Five trait scores. We then prompt the model with these descriptions and questions from the Alpaca questionnaire, allowing us to sample hidden activations that vary along personality traits in known, quantifiable ways. Using linear regression, we learn a set of per-layer directions in activation space, and test their effectiveness for probing and steering model behavior. Our results suggest that linear directions aligned with trait-scores are effective probes for personality detection, while their steering capabilities strongly depend on context, producing reliable effects in forced-choice tasks but limited influence in open-ended generation or when additional context is present in the prompt.
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。
