Retrieval augmented generation in multimodal and text-only settings.
We are developing language models that can process any written language.
Design, understand, and evaluate multimodality in multilingual models