Machine Learning Interatomic Potentials (MLIPs) have positioned themselves as a key tool for atomistic modeling in materials science. MLIPs cover an expansive range of systems, taking advantage of the highly accurate electronic structure calculations based on quantum mechanics, but at a significantly lower computational cost. They allow to scale up atomistic simulations to larger systems, longer timescales, and more complex phenomena; they therefore significantly contribute to the acceleration of the discovery of novel structural and functional materials, and in the advancements in our understanding of matter. Ground-breaking bodies of work have been published since the seminal work of Behler and Parrinello in 2007 [1], transforming the field into a rapidly evolving research discipline [2-16]. However, alongside these advancements, a crucial challenge emerges: the need for standardized protocols for MLIP generation and storage, as well as comprehensive, accessible databases for ab initio datasets. This question is very much at the heart of the French national program "DIADEM" and its digital infrastructure "DIAMOND", which sponsor in part this workshop.

Location: amphithéâtre Besson, campus Phelma, Batiment A, 1130 rue de la Piscine, 38400 Saint Martin d'Hères. Campus map here (look for building A, Phelma campus): https://dropsu.sorbonne-universite.fr/s/gdMJ2o5FZtKn48Z

How to reach the event? The airport shuttle will drop you at the Grenoble train station. From there, take tram B (towards "Gières, Plaine des Sports"), and stop at "Bibliothèques Universitaires". Walk north and reach Building A on the map.

Follow remotely the presentations:

https://grenoble-inp.zoom.us/j/92354659113
Meeting ID: 923 5465 9113
Password: 968905