We are developing language models that can process any written language.
Retrieval augmentation of multimodal models for image captioning.
Design, understand, and evaluate multimodality in multilingual models