DVPS

Artificial intelligence today is confined to the representation of the world found in datasets. We’re building a new class of AI systems designed to gain a grounded understanding through direct interaction with the physical world. DVPS stands for Diversibus Viis Plurima Solvo, Latin for “Through diverse paths, I solve many issues”. This is one of the most heavily backed projects in the Horizon Europe program, bringing together 20 leading organizations across 9 countries to advance the science and engineering of multimodal foundation models (MMFMs). We are developing an open-source toolkit to streamline the design, pre-training, fine-tuning, and modality expansion of MMFMs. It supports the reuse and composition of pre-trained models, reducing development time and cost while enhancing adaptability across modalities. DVPS is pushing this frontier forward by driving research and innovation in multimodal foundation models (MMFMs). These AI systems can reason and operate across text, images, audio, and more.