SCAI was pleased to take part in the inauguration of two new joint research teams in Artificial Intelligence: ARCHES and MEGAVOLT. Developed within the laboratories of Sorbonne University in collaboration with UVSQ – Université Paris-Saclay, CNRS and Inria, these initiatives further strengthen the excellence of AI research at the university.
Presented by Claire Monteleoni, project leader, together with Cécile Mallet, ARCHES aims to enhance the scientific understanding of the climate system by leveraging the latest advancements in Artificial Intelligence.
The team focuses on supporting society’s response to climate change — improving prediction capabilities, contributing to mitigation strategies, and optimising adaptation efforts.
The presentation was followed by a talk from Emmanuel de Bézenac, who discussed efficient construction of generative models for weather forecasting - an innovative approach bridging AI and atmospheric sciences.
MEGAVOLT — MachinE learninG et équAtions d’éVOLuTion — was introduced by Bruno Després, head of the team.
Its goal is to build strong connections between rigorous mathematical modelling (particularly evolution equations) and the predictive power of machine learning, paving the way for new hybrid models that are both robust and interpretable.
During the session, Gérard Biau, member of MEGAVOLT and Director of SCAI, also presented the mission and vision of the cluster: to bring together AI expertise across Sorbonne University and accelerate interdisciplinary collaboration.
The session concluded with a scientific talk by Raphaël Berthier entitled “Theory of some simple neural networks”, highlighting progress in foundational learning theory.
A vibrant research community in action
A poster session followed the presentations, offering the opportunity for enriching discussions with members of both teams.
The contributions showcased the strong commitment of researchers to tackling ambitious and interdisciplinary scientific challenges.