DYNAMO Inria Associate Team

DYNamical systems, Analysis, and Machine learning for self-Organization of matter

dynamo.png

Inria Associate Team

MALICE (France)

CSML (Italy)

DYNAMO is an Inria Associate Team jointly established between the MALICE Inria project-team at Université Jean Monnet (Saint-Étienne, France) and the Computational Statistics and Machine Learning (CSML) group at the Istituto Italiano di Tecnologia (Genoa, Italy).

It brings together researchers from France and Italy to explore the intersection of machine learning, dynamical systems, and physics-informed modeling.

The team aims to develop new mathematical and computational frameworks to learn and control self-organizing phenomena—complex processes through which matter spontaneously forms ordered patterns under laser excitation. These efforts address fundamental scientific questions while enabling practical advances in surface engineering, with potential applications spanning materials science, energy, health, and nanotechnology.

news

Nov 01, 2025 We will be at the EurIPS 2025 workshop SysDiff to present our work on Douglas-Rachford Splitting for Hybrid Differentiable Models.
Sep 18, 2025 Our paper Conformal Online Learning of Deep Koopman Linear Embeddings has been accepted to NeurIPS 2025!
Mar 20, 2025 Creation of the Inria Associate Team DYNAMO.