Synthetic pre-training for neural-network interatomic potentials

Machine learning (ML) based interatomic potentials have transformed the field of atomistic materials modelling. However, ML potentials depend critically on the quality and quantity of quantum-mechanical reference data with which they are trained, and therefore developing datasets and training pipeli...

Full description

Saved in:
Bibliographic Details
Published in:Machine learning: science and technology 2024-03, Vol.5 (1), p.15003
Main Authors: Gardner, John L A, Baker, Kathryn T, Deringer, Volker L
Format: Article
Language:English
Subjects:
atomistic simulations
Datasets
Machine learning
molecular dynamics
Neural networks
Synthetic data
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Machine learning (ML) based interatomic potentials have transformed the field of atomistic materials modelling. However, ML potentials depend critically on the quality and quantity of quantum-mechanical reference data with which they are trained, and therefore developing datasets and training pipelines is becoming an increasingly central challenge. Leveraging the idea of ‘synthetic’ (artificial) data that is common in other areas of ML research, we here show that synthetic atomistic data, themselves obtained at scale with an existing ML potential, constitute a useful pre-training task for neural-network (NN) interatomic potential models. Once pre-trained with a large synthetic dataset, these models can be fine-tuned on a much smaller, quantum-mechanical one, improving numerical accuracy and stability in computational practice. We demonstrate feasibility for a series of equivariant graph-NN potentials for carbon, and we carry out initial experiments to test the limits of the approach.
ISSN:2632-2153
2632-2153
DOI:10.1088/2632-2153/ad1626