Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | https://arxiv.org/abs/2605.06115 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866917470878564352 |
|---|---|
| author | Zeng, Zhen Gu, Leijiang Li, Feng Yu, Jing Shi, Zenglin |
| author_facet | Zeng, Zhen Gu, Leijiang Li, Feng Yu, Jing Shi, Zenglin |
| contents | Multimodal Large Language Models (MLLMs), trained primarily on English-centric data, frequently generate culturally inappropriate or misaligned responses in cross-cultural settings. To mitigate this, we introduce the task of cross-cultural knowledge insertion, which focuses on adapting models to specific cultural contexts while preserving their original behavior in other cultures. To facilitate research in this area, we introduce CrossCult-KIBench, a comprehensive evaluation benchmark for assessing both the effectiveness of knowledge insertion and its unintended side effects on non-target cultures. The benchmark includes 9,800 image-grounded cases covering 49 culturally relevant visual scenarios across English, Chinese, and Arabic language-culture groups. It supports evaluation in both single-insert and sequential-insert settings. We also propose Memory-Conditioned Knowledge Insertion (MCKI) as a baseline method. MCKI retrieves relevant cultural knowledge from an external memory using frozen MLLM representations, prepending matched entries as conditional prompts when applicable. Extensive experiments on CrossCult-KIBench reveal that current approaches struggle to balance effective cultural adaptation with behavioral preservation, highlighting a key challenge in developing culturally-aware MLLMs. Our work thus underscores an important research direction for developing more culturally adaptive and responsible MLLMs. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_06115 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | CrossCult-KIBench: A Benchmark for Cross-Cultural Knowledge Insertion in MLLMs Zeng, Zhen Gu, Leijiang Li, Feng Yu, Jing Shi, Zenglin Artificial Intelligence Multimodal Large Language Models (MLLMs), trained primarily on English-centric data, frequently generate culturally inappropriate or misaligned responses in cross-cultural settings. To mitigate this, we introduce the task of cross-cultural knowledge insertion, which focuses on adapting models to specific cultural contexts while preserving their original behavior in other cultures. To facilitate research in this area, we introduce CrossCult-KIBench, a comprehensive evaluation benchmark for assessing both the effectiveness of knowledge insertion and its unintended side effects on non-target cultures. The benchmark includes 9,800 image-grounded cases covering 49 culturally relevant visual scenarios across English, Chinese, and Arabic language-culture groups. It supports evaluation in both single-insert and sequential-insert settings. We also propose Memory-Conditioned Knowledge Insertion (MCKI) as a baseline method. MCKI retrieves relevant cultural knowledge from an external memory using frozen MLLM representations, prepending matched entries as conditional prompts when applicable. Extensive experiments on CrossCult-KIBench reveal that current approaches struggle to balance effective cultural adaptation with behavioral preservation, highlighting a key challenge in developing culturally-aware MLLMs. Our work thus underscores an important research direction for developing more culturally adaptive and responsible MLLMs. |
| title | CrossCult-KIBench: A Benchmark for Cross-Cultural Knowledge Insertion in MLLMs |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2605.06115 |