Benchmarking machine-learning interatomic potentials using classical force fields
Abstract
We propose a rational comparison between several well-known machine-learning interatomic potentials (linear atomic cluster expansion, the Behler-Parrinello high-dimensional neural network potential, and the graph neural network-based MACE potential) by testing their performance for fitting classical potentials such as the Lennard-Jones and Stillinger-Weber potential. With this we assess their advantages and limitations regarding several aspects: varying training dataset size, robustness to noise, carbon footprint, as well as the ability to reproduce the radial distribution function as an example observable. Overall the MACE models are shown to most accurately reproduce the potential energy surface, with little sign of overfitting, but at a significant computational cost, highlighting important questions regarding the choice of models in computational materials science.
Related
- Benchmarking Universal Machine-Learning Interatomic Potentials for Temperature-Dependent Elasticity of Binary and High-Entropy Refractory Carbides
- Benchmarking Machine-Learning Interatomic Potentials for Dynamical Stability in Inorganic Semiconductor Nanocrystals: A CdSe Case Study
- MLIP Detective: Active Failure Mode Discovery Beyond Benchmark Scores for Machine-Learning Interatomic Potentials
- Cluster-based Structural Similarity for Dataset Visualization and Data Selection for Machine Learning Interatomic Potentials
- Leveraging Machine Learning Interatomic Potentials into Fitting Fast and Accurate AMBER-Compatible Precise Force Field
BibTeX
@misc{sandberg2026benchmarking,
title = {{Benchmarking machine-learning interatomic potentials using classical force fields}},
author = {Johannes Sandberg and Geneviève Dusson and Julien Lam},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.14239},
eprint = {2610.14239},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.14239},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Benchmarking machine-learning interatomic potentials using classical force fields
AU - Johannes Sandberg
AU - Geneviève Dusson
AU - Julien Lam
PY - 2026
DA - 2026/10/09
PB - arXiv
DO - 10.48550/arxiv.2610.14239
UR - https://arxiv.org/abs/2610.14239
AB - We propose a rational comparison between several well-known machine-learning interatomic potentials (linear atomic cluster expansion, the Behler-Parrinello high-dimensional neural network potential, and the graph neural network-based MACE potential) by testing their performance for fitting classical potentials such as the Lennard-Jones and Stillinger-Weber potential. With this we assess their advantages and limitations regarding several aspects: varying training dataset size, robustness to noise, carbon footprint, as well as the ability to reproduce the radial distribution function as an example observable. Overall the MACE models are shown to most accurately reproduce the potential energy surface, with little sign of overfitting, but at a significant computational cost, highlighting important questions regarding the choice of models in computational materials science.
ER -