From Surveys to Narratives: Rethinking Cultural Value Adaptation in LLMs
This addresses cultural bias in LLMs for users in diverse applications, but it is incremental as it builds on prior survey-based methods.
The paper tackled adapting cultural values in LLMs by showing that relying solely on World Values Survey data homogenizes norms and interferes with factual knowledge, and augmenting it with narratives from Wikipedia and NormAd consistently improves cultural distinctiveness.
Adapting cultural values in Large Language Models (LLMs) presents significant challenges, particularly due to biases and limited training data. Prior work primarily aligns LLMs with different cultural values using World Values Survey (WVS) data. However, it remains unclear whether this approach effectively captures cultural nuances or produces distinct cultural representations for various downstream tasks. In this paper, we systematically investigate WVS-based training for cultural value adaptation and find that relying solely on survey data can homogenize cultural norms and interfere with factual knowledge. To investigate these issues, we augment WVS with encyclopedic and scenario-based cultural narratives from Wikipedia and NormAd. While these narratives may have variable effects on downstream tasks, they consistently improve cultural distinctiveness than survey data alone. Our work highlights the inherent complexity of aligning cultural values with the goal of guiding task-specific behavior. We release our code at https://github.com/faridlazuarda/from-surveys-to-narratives.