读懂提示词背后的含义：刻板印象如何塑造 LLM 的隐式个性化
Reading Between the Prompts: How Stereotypes Shape LLM's Implicit Personalization
AI总结：
本研究探讨了 LLMs 如何基于对话中的刻板印象线索进行隐式个性化，导致对少数群体的质量下降，并提出利用线性探针干预模型内部表征以有效缓解该问题，强调需提升用户身份表征的透明度与控制力。
摘要：
生成式大型语言模型（LLMs）能从对话中的细微线索推断用户的人口统计信息——这一现象被称为隐式个性化。先前的研究表明，即使未明确提供人口统计信息，此类推断也可能导致假定的少数群体用户获得较低质量的回复。在本研究中，我们使用受控合成对话，系统探讨 LLMs 如何响应刻板印象线索，通过模型内部机制和针对用户问题的生成答案来分析模型的潜在用户表征。研究结果表明，LLMs 确实会基于这些刻板印象信号推断人口统计属性；对于某些群体，即使用户明确表明自己属于不同的人口群体，这种推断依然存在。最后，我们证明，通过使用训练好的线性探针干预模型的内部表征，引导其向明确声明的身份靠拢，可以有效缓解这种由刻板印象驱动的隐式个性化。我们的结果强调了在 LLMs 表征用户身份方面需要更高的透明度与控制力。
英文摘要：
Generative Large Language Models (LLMs) infer user's demographic information from subtle cues in the conversation -- a phenomenon called implicit personalization. Prior work has shown that such inferences can lead to lower quality responses for users assumed to be from minority groups, even when no demographic information is explicitly provided. In this work, we systematically explore how LLMs respond to stereotypical cues using controlled synthetic conversations, by analyzing the models' latent user representations through both model internals and generated answers to targeted user questions. Our findings reveal that LLMs do infer demographic attributes based on these stereotypical signals, which for a number of groups even persists when the user explicitly identifies with a different demographic group. Finally, we show that this form of stereotype-driven implicit personalization can be effectively mitigated by intervening on the model's internal representations using a trained linear probe to steer them toward the explicitly stated identity. Our results highlight the need for greater transparency and control in how LLMs represent user identity.
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。
