The birth of Foundation Models brought unprecedented results in a wide range of tasks, from language to vision, to robotic control. These models are able to process huge quantities of data, and can extract and develop rich representations, which can be employed across different domains and modalities. However, they still have issues in adapting to dynamic, real-world scenarios without retraining the entire model from scratch. In this work, we propose the application of Continual Learning and Compositionality principles to foster the development of more flexible, efficient and smart AI solutions.
@article{arxiv.2510.18608,
title = {A Compositional Paradigm for Foundation Models: Towards Smarter Robotic Agents},
author = {Luigi Quarantiello and Elia Piccoli and Jack Bell and Malio Li and Giacomo Carfì and Eric Nuertey Coleman and Gerlando Gramaglia and Lanpei Li and Mauro Madeddu and Irene Testa and Vincenzo Lomonaco},
journal= {arXiv preprint arXiv:2510.18608},
year = {2025}
}