CompChem Observer

Conferences, workshops and schools in computational chemistry

Automating atomistic machine learning

Dates
Type
workshop
Format
in-person
Location
CECAM-HQ-EPFL, Lausanne, Switzerland
Organiser
CECAM
Official page
https://www.cecam.org/workshop-details/automating-atomistic-machine-learning-1494
Topics
ml-potentialssoftware-hpcmaterials-modeling

This workshop focuses on automating machine-learned interatomic potentials (MLIPs) to make them a routine, accessible tool in computational materials modelling. It covers integrating MLIPs into automation frameworks, tackling challenges like active learning, GPU training, and data provenance. Topics include automating data selection, descriptors, and regression. Key discussions address which parts of MLIP development should be automated, interfacing with DFT infrastructure, the role of pre-trained foundation models and AI agents, and future applications.

Deadlines

Deadlines as published by the organiser.
DeadlineDateTimezoneNote
ApplicationAoE (Anywhere on Earth)
RegistrationAoE (Anywhere on Earth)

Suggest a correction ·Report this event