The Curious Case of Curiosity across Human Cultures and LLMs

2025-10-14

AI Summary

This study investigates whether large language models (LLMs) accurately represent cross-cultural differences in curiosity—particularly whether they systematically exhibit Western-centric biases in expression. Method: Leveraging a multilingual Yahoo! Answers corpus, we introduce CUEST, the first cross-societal curiosity evaluation framework, integrating stylistic and thematic preference analysis, social-scientific construct modeling, and fine-tuning techniques for both open- and closed-weight LLMs. Contribution/Results: Empirical results show that mainstream LLMs significantly attenuate non-Western cultural markers, converging toward Western curiosity expressions. Cultural-aware fine-tuning improves alignment between human and model curiosity representations by up to 50%. This work provides the first quantitative characterization of cultural bias in LLMs with respect to curiosity—a core cognitive dimension—and demonstrates both its measurable impact on cross-cultural adaptability and its tractability through targeted optimization.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Cognitive Modeling & Cognitive Systems: Computational Creativity

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Recent advances in Large Language Models (LLMs) have expanded their role in human interaction, yet curiosity -- a central driver of inquiry -- remains underexplored in these systems, particularly across cultural contexts. In this work, we investigate cultural variation in curiosity using Yahoo! Answers, a real-world multi-country dataset spanning diverse topics. We introduce CUEST (CUriosity Evaluation across SocieTies), an evaluation framework that measures human-model alignment in curiosity through linguistic (style), topic preference (content) analysis and grounding insights in social science constructs. Across open- and closed-source models, we find that LLMs flatten cross-cultural diversity, aligning more closely with how curiosity is expressed in Western countries. We then explore fine-tuning strategies to induce curiosity in LLMs, narrowing the human-model alignment gap by up to 50%. Finally, we demonstrate the practical value of curiosity for LLM adaptability across cultures, showing its importance for future NLP research.

Problem

• Investigating cultural variations in curiosity across human societies and LLMs
• Developing evaluation framework to measure human-model curiosity alignment
• Addressing LLMs' Western bias in curiosity expression across cultures

Innovation

• CUEST framework evaluates cross-cultural curiosity alignment
• Fine-tuning strategies narrow human-model curiosity gap
• Linguistic and topic analysis measures cultural curiosity variation