{"ID":2851833,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.21849","arxiv_id":"2510.21849","title":"TowerVision: Understanding and Improving Multilinguality in Vision-Language Models","abstract":"Despite significant advances in vision-language models (VLMs), most existing work follows an English-centric design process, limiting their effectiveness in multilingual settings. In this work, we provide a comprehensive empirical study analyzing the impact of several multilingual design choices, such as training data composition, encoder selection, and text backbones. The result is TowerVision, a family of open multilingual VLMs for both image-text and video-text tasks, built upon the multilingual text-only model Tower+. TowerVision achieves competitive performance on multiple multimodal multilingual benchmarks and shows particular strength in culturally grounded tasks and multimodal translation. By incorporating visual and cultural context during fine-tuning, our models surpass existing approaches trained on substantially larger datasets, as demonstrated on ALM-Bench and Multi30K (image tasks) and ViMUL-Bench (video tasks). Alongside the models, we release VisionBlocks, a high-quality, curated vision-language dataset. Our findings highlight that multilingual vision-language training data substantially improves cross-lingual generalization -- both from high-resource to underrepresented languages and vice versa -- and that instruction-tuned LLMs are not always the optimal initialization point. To support further research, we publicly release all models, data, and training recipes.","short_abstract":"Despite significant advances in vision-language models (VLMs), most existing work follows an English-centric design process, limiting their effectiveness in multilingual settings. In this work, we provide a comprehensive empirical study analyzing the impact of several multilingual design choices, such as training data...","url_abs":"https://arxiv.org/abs/2510.21849","url_pdf":"https://arxiv.org/pdf/2510.21849v3","authors":"[\"André G. Viveiros\",\"Patrick Fernandes\",\"Saul Santos\",\"Sonal Sannigrahi\",\"Emmanouil Zaranis\",\"Nuno M. Guerreiro\",\"Amin Farajian\",\"Pierre Colombo\",\"Graham Neubig\",\"André F. T. Martins\"]","published":"2025-10-22T17:02:48Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"cs.AI\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false}