Who Laughs with Whom? Disentangling Influential Factors in Humor Preferences across User Clusters and LLMs

TL;DR

This paper investigates how humor preferences differ across user groups and how large language models can be guided to align with specific humor tastes using clustering and persona prompting.

Contribution

It introduces a method to model and analyze heterogeneity in humor preferences across user clusters and demonstrates how LLMs can be directed to match these preferences.

Findings

User clusters show distinct humor preference patterns.

LLMs can mimic specific user cluster preferences.

Persona prompting can steer LLM humor responses.

Abstract

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.