ValuesRAG: Enhancing Cultural Alignment Through Retrieval-Augmented Contextual Learning
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866913825936113664 |
|---|---|
| author | Seo, Wonduk Yuan, Zonghao Bu, Yi |
| author_facet | Seo, Wonduk Yuan, Zonghao Bu, Yi |
| contents | Ensuring cultural values alignment in Large Language Models (LLMs) remains a critical challenge, as these models often embed Western-centric biases from their training data, leading to misrepresentations and fairness concerns in cross-cultural applications. Existing approaches such as role assignment and few-shot learning struggle to address these limitations effectively due to their reliance on pre-trained knowledge, limited scalability, and inability to capture nuanced cultural values. To address these issues, we propose ValuesRAG, a novel and effective framework that applies Retrieval-Augmented Generation (RAG) with In-Context Learning (ICL) to integrate cultural and demographic knowledge dynamically during text generation. Leveraging the World Values Survey (WVS) dataset, ValuesRAG first generates summaries of values for each individual. We subsequently curate several representative regional datasets to serve as test datasets and retrieve relevant summaries of values based on demographic features, followed by a reranking step to select the top-k relevant summaries. We evaluate ValuesRAG using 6 diverse regional datasets and show that it consistently outperforms baselines: including zero-shot, role-assignment, few-shot, and hybrid methods, both in main experiments and ablation settings. Notably, ValuesRAG achieves the best overall performance over prior methods, demonstrating its effectiveness in fostering culturally aligned and inclusive AI systems. Our findings underscore the potential of dynamic retrieval-based methods to bridge the gap between global LLM capabilities and localized cultural values. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_01031 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | ValuesRAG: Enhancing Cultural Alignment Through Retrieval-Augmented Contextual Learning Seo, Wonduk Yuan, Zonghao Bu, Yi Computation and Language Artificial Intelligence Social and Information Networks Ensuring cultural values alignment in Large Language Models (LLMs) remains a critical challenge, as these models often embed Western-centric biases from their training data, leading to misrepresentations and fairness concerns in cross-cultural applications. Existing approaches such as role assignment and few-shot learning struggle to address these limitations effectively due to their reliance on pre-trained knowledge, limited scalability, and inability to capture nuanced cultural values. To address these issues, we propose ValuesRAG, a novel and effective framework that applies Retrieval-Augmented Generation (RAG) with In-Context Learning (ICL) to integrate cultural and demographic knowledge dynamically during text generation. Leveraging the World Values Survey (WVS) dataset, ValuesRAG first generates summaries of values for each individual. We subsequently curate several representative regional datasets to serve as test datasets and retrieve relevant summaries of values based on demographic features, followed by a reranking step to select the top-k relevant summaries. We evaluate ValuesRAG using 6 diverse regional datasets and show that it consistently outperforms baselines: including zero-shot, role-assignment, few-shot, and hybrid methods, both in main experiments and ablation settings. Notably, ValuesRAG achieves the best overall performance over prior methods, demonstrating its effectiveness in fostering culturally aligned and inclusive AI systems. Our findings underscore the potential of dynamic retrieval-based methods to bridge the gap between global LLM capabilities and localized cultural values. |
| title | ValuesRAG: Enhancing Cultural Alignment Through Retrieval-Augmented Contextual Learning |
| topic | Computation and Language Artificial Intelligence Social and Information Networks |
| url | https://arxiv.org/abs/2501.01031 |