Regularity priors for the linear atomic cluster expansion
Open Access
- 23 March 2026
- journal article
- Published by IOP Publishing in Machine Learning: Science and Technology
- Vol. 7 (2)
Abstract
Machine-learned interatomic potentials enable large systems to be simulated for long time scales at near ab-initio accuracy. This accuracy is achieved by fitting extremely flexible model architectures to high quality reference data. In practice, this flexibility can cause unwanted behavior such as jagged predicted potential energy surfaces and generally poor out-of-distribution behavior. We investigate a general strategy for incorporating prior beliefs on the regularity of the target energy into linear ACE models and explore to what extent this approach improves the quality of the fitted models. Our main focus is an over-regularisation that replicates the Gaussian broadening used in SOAP descriptors within the ACE framework. Numerical tests indicate that the exact form of the prior is non-critical but that including such a prior leads to significant improvement in test errors, consistent repulsion at close-approach, eliminates spurious false minima in the potential energy and enhances stability during molecular dynamics simulations.Keywords
Funding Information
- Natural Sciences and Engineering Research Council of Canada
- EPSRC Centre for Doctoral Training (EP/S023828/1)
- CASTEP-USER
- UK Research and Innovation (EP/W030438/1)
- Engineering and Physical Sciences Research Council (EP/X035891/1)
This publication has 52 references indexed in Scilit:
- Atomic cluster expansion without self-interactionJournal of Computational Physics, 2024
- Scalable Parallel Algorithm for Graph Neural Network Interatomic Potentials in Molecular Dynamics SimulationsJournal of Chemical Theory and Computation, 2024
- Atomic cluster expansion: Completeness, efficiency and stabilityJournal of Computational Physics, 2022
- Gaussian Process Regression for Materials and MoleculesChemical Reviews, 2021
- Gibbs Ensemble Monte Carlo for Reactive Force Fields to Determine the Vapor–Liquid Equilibrium of CO2 and H2OJournal of Chemical Theory and Computation, 2020
- The ORCA program systemWIREs Computational Molecular Science, 2011
- Second-order integrators for Langevin equations with holonomic constraintsChemical Physics Letters, 2006
- Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis setComputational Materials Science, 1996
- On the limited memory BFGS method for large scale optimizationMathematical Programming, 1989
- The Stopping and Range of Ions in Matter1985