分析用户和大模型的笑点偏好差异，发现可定向调整模型幽默风格。
Who Laughs with Whom? Disentangling Influential Factors in Humor Preferences across User Clusters and LLMs
- 按用户投票行为聚类，用布鲁达-泰勒-卢斯模型解耦幽默偏好因子。
- 不同用户群偏好模式各异，部分大模型表现类似特定用户群。
- 通过角色提示可引导大模型模仿特定群体的幽默偏好。
幽默偏好在个体与文化间差异显著，给大语言模型（LLMs）评估幽默带来挑战。本研究基于日本创意回应游戏Oogiri的用户投票日志，通过聚类分析用户，并利用布鲁达-泰勒-卢斯（Bradley-Terry-Luce）模型估计各用户群对可解释幽默偏好因子的专属权重。我们通过提示让大模型判断哪个回答更有趣，发现用户群呈现明显不同的偏好模式，且部分大模型结果与特定用户群高度相似。最后，我们证明通过角色提示（persona prompting），可引导大模型偏好向特定用户群靠拢。数据采集与分析脚本将公开，以支持可复现性。
原文摘要 · Abstract (English)
Humor preferences vary widely across individuals and cultures, complicating the evaluation of humor using large language models (LLMs). In this study, we model heterogeneity in humor preferences in Oogiri, a Japanese creative response game, by clustering users with voting logs and estimating cluster-specific weights over interpretable preference factors using Bradley-Terry-Luce models. We elicit preference judgments from LLMs by prompting them to select the funnier response and found that user clusters exhibit distinct preference patterns and that the LLM results can resemble those of particular clusters. Finally, we demonstrate that, by persona prompting, LLM preferences can be directed toward a specific cluster. The scripts for data collection and analysis will be released to support reproducibility.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。