Development of Continual Fine-tuning Methodologies for Foundation Models
Scientific-Disciplinary Group
01/INFO-01 - Informatics
Description
The project aims to develop a modular and decentralized software library for Collective Artificial Intelligence (AI) methodologies, fostering cooperation, reuse, and sustainability. Based on the COLLAGE approach—centered on compositionality and reusability—it seeks to build an infrastructure enabling the incremental and distributed creation, combination, and adaptation of AI solutions. The library will serve as a platform for integrating interoperable intelligent components and protocols for collective learning among models and applications. COLLAGE introduces a new paradigm for sustainable and cooperative AI, validated through a proof of concept in a decentralized system of “collective AI for flood prediction” in Europe.
Compensation
33,725 Euro
Job posting website
Number of positions
1
Maximum duration
24.0
Funding body
Luiss Guido Carli
View the original posting on the MUR website: Go to MUR website