Synthetic pre-training for neural-network interatomic potentials
Open Access
- 10 January 2024
- journal article
- research article
- Published by IOP Publishing in Machine Learning: Science and Technology
- Vol. 5 (1) , 015003
Abstract
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.Keywords
All Related Versions
Funding Information
- Engineering and Physical Sciences Research Council (EP/T517811/1)
- UK Research and Innovation (EP/X016188/1)
This publication has 52 references indexed in Scilit:
- Atomic Cluster Expansion for Quantum-Accurate Large-Scale Simulations of CarbonJournal of Chemical Theory and Computation, 2023
- Improving molecular property prediction through a task similarity enhanced transfer learning strategyiScience, 2022
- LAMMPS - a flexible simulation tool for particle-based materials modeling at the atomic, meso, and continuum scalesComputer Physics Communications, 2022
- Cluster Fragments in Amorphous Phosphorus and their Evolution under PressureAdvanced Materials, 2021
- Machine Learning Force FieldsChemical Reviews, 2021
- Machine Learning Interatomic Potentials as Emerging Tools for Materials ScienceAdvanced Materials, 2019
- DeePMD-kit: A deep learning package for many-body potential energy representation and molecular dynamicsComputer Physics Communications, 2018
- First Principles Neural Network Potentials for Reactive Simulations of Large Molecular and Condensed SystemsAngewandte Chemie International Edition, 2017
- Graphitization of amorphous carbons: A comparative study of interatomic potentialsCarbon, 2016
- The potential of atomistic simulations and the knowledgebase of interatomic modelsJOM, 2011