CALM: Culturally Self-Aware Language Models

TL;DR

CALM introduces a framework that enhances language models with dynamic cultural self-awareness by disentangling, structuring, and adaptively integrating cultural signals, leading to improved performance on cross-cultural tasks.

Contribution

This work presents CALM, a novel approach that models cultural signals explicitly and adaptively within language models, enabling genuine cultural sensitivity and continual self-improvement.

Findings

CALM outperforms existing methods on multiple cross-cultural benchmarks.

The framework effectively disentangles task semantics from cultural concepts.

Cultural clusters improve the model's contextual understanding.

Abstract

Cultural awareness in language models is the capacity to understand and adapt to diverse cultural contexts. However, most existing approaches treat culture as static background knowledge, overlooking its dynamic and evolving nature. This limitation reduces their reliability in downstream tasks that demand genuine cultural sensitivity. In this work, we introduce CALM, a novel framework designed to endow language models with cultural self-awareness. CALM disentangles task semantics from explicit cultural concepts and latent cultural signals, shaping them into structured cultural clusters through contrastive learning. These clusters are then aligned via cross-attention to establish fine-grained interactions among related cultural features and are adaptively integrated through a Mixture-of-Experts mechanism along culture-specific dimensions. The resulting unified representation is fused with the model's original knowledge to construct a culturally grounded internal identity state, which is further enhanced through self-prompted reflective learning, enabling continual adaptation and self-correction. Extensive experiments conducted on multiple cross-cultural benchmark datasets demonstrate that CALM consistently outperforms state-of-the-art methods.