大型语言模型会为景观付费吗？从主观选择中推断支付意愿
Would a Large Language Model Pay Extra for a View? Inferring Willingness to Pay from Subjective Choices
AI总结：
研究在旅行助手场景下，通过多分类逻辑模型分析LLM的主观选择，推断其支付意愿并与人类基准比较，发现LLM在属性层面存在系统偏差且高估支付意愿，但通过条件化偏好可改善。
摘要：
随着大型语言模型（LLM）越来越多地部署在旅行辅助和购买支持等应用中，它们常常需要在没有客观正确答案的情况下代表用户做出主观选择。我们在旅行助手背景下研究LLM的决策，通过向模型呈现选择困境，并使用多项逻辑模型分析其响应，推导出隐含的支付意愿（WTP）估计。随后将这些WTP值与经济学文献中的人类基准值进行比较。除了基线设置外，我们还研究了在更现实条件下模型行为的变化，包括提供用户过去选择的信息和基于角色的提示。我们的结果表明，虽然可以从较大的LLM中推导出有意义的WTP值，但它们在属性层面也显示出系统偏差。此外，它们倾向于整体高估人类的WTP，特别是在引入昂贵选项或面向商业的角色时。将模型条件化于对更便宜选项的先前偏好，得出的估值更接近人类基准。总体而言，我们的发现突出了使用LLM进行主观决策支持的潜力和局限性，并强调了在实际部署此类系统时仔细选择模型、设计提示和表示用户的重要性。
英文摘要：
As Large Language Models (LLMs) are increasingly deployed in applications such as travel assistance and purchasing support, they are often required to make subjective choices on behalf of users in settings where no objectively correct answer exists. We study LLM decision-making in a travel-assistant context by presenting models with choice dilemmas and analyzing their responses using multinomial logit models to derive implied willingness to pay (WTP) estimates. These WTP values are subsequently compared to human benchmark values from the economics literature. In addition to a baseline setting, we examine how model behavior changes under more realistic conditions, including the provision of information about users' past choices and persona-based prompting. Our results show that while meaningful WTP values can be derived for larger LLMs, they also display systematic deviations at the attribute level. Additionally, they tend to overestimate human WTP overall, particularly when expensive options or business-oriented personas are introduced. Conditioning models on prior preferences for cheaper options yields valuations that are closer to human benchmarks. Overall, our findings highlight both the potential and the limitations of using LLMs for subjective decision support and underscore the importance of careful model selection, prompt design, and user representation when deploying such systems in practice.
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。
