The project is in an experimental, pre-alpha, exploratory phase with the intention to be productionized. We move fast, break things, and explore various aspects of the seamless developer experience ...

Abstract: Visual grounding seeks to localize the image region corresponding to a free-form text description. Recently, the strong multimodal capabilities of Large Vision-Language Models (LVLMs) have ...

Abstract: Programming based approaches to reasoning tasks have substantially expanded the types of questions models can answer about visual scenes. Yet on benchmark visual reasoning data, when models ...