规则与现实之间：关于大语言模型道德判断的上下文敏感性
On the Context Sensitivity of LLM Moral Judgment
AI总结：
研究探讨大语言模型在不同上下文中的道德判断敏感性，通过引入上下文道德选择数据集，发现多数模型对上下文变化敏感，且与人类判断存在差异，提出激活引导方法以控制上下文敏感性。
摘要：
人类的道德决策严重依赖于上下文。然而，关于LLM道德性的研究大多集中在固定场景。我们通过引入上下文道德选择数据集，该数据集包含经过道德心理学验证的道德困境，系统性地变化上下文以转移人类判断：结果导向、情感和关系。评估22个LLM后，发现几乎所有模型对上下文敏感，其判断倾向于违反规则的行为。与人类调查比较发现，模型和人类对不同上下文变化的反应不同，且在基础情况下与人类判断一致的模型在上下文敏感性上未必一致。这引发了对控制上下文敏感性的疑问，我们通过激活引导方法可靠地增加或减少模型的上下文敏感性。
英文摘要：
A human's moral decision depends heavily on the context. Yet research on LLM morality has largely studied fixed scenarios. We address this gap by introducing Contextual MoralChoice, a dataset of moral dilemmas with systematic contextual variations known from moral psychology to shift human judgment: consequentialist, emotional, and relational. Evaluating 22 LLMs, we find that nearly all models are context-sensitive, shifting their judgments towards rule-violating behavior. Comparing with a human survey, we find that models and humans are most triggered by different contextual variations, and that a model aligned with human judgments in the base case is not necessarily aligned in its contextual sensitivity. This raises the question of controlling contextual sensitivity, which we address with an activation steering approach that can reliably increase or decrease a model's contextual sensitivity. Code and data: https://github.com/adrian-sauter/contextual_moralchoice.
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。
