Saved in:
Bibliographic Details
Main Authors: Kuryla, Domantas, Berger, Fabian, Csányi, Gábor, Michaelides, Angelos
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2510.19774
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909864142307328
author Kuryla, Domantas
Berger, Fabian
Csányi, Gábor
Michaelides, Angelos
author_facet Kuryla, Domantas
Berger, Fabian
Csányi, Gábor
Michaelides, Angelos
contents Training of general-purpose machine learning interatomic potentials (MLIPs) relies on large datasets with properties usually computed with density functional theory (DFT). A pre-requisite for accurate MLIPs is that the DFT data are well converged to minimize numerical errors. A possible symptom of errors in DFT force components is nonzero net force. Here, we consider net forces in datasets including SPICE, Transition1x, ANI-1x, ANI-1xbb, AIMNet2, QCML, and OMol25. Several of these datasets suffer from significant nonzero DFT net forces. We also quantify individual force component errors by comparison to recomputed forces using more reliable DFT settings at the same level of theory, and we find significant discrepancies in force components averaging from 1.7 meV/Å in the SPICE dataset to 33.2 meV/Å in the ANI-1x dataset. These findings underscore the importance of well converged DFT data as increasingly accurate MLIP architectures become available.
format Preprint
id arxiv_https___arxiv_org_abs_2510_19774
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle How Accurate Are DFT Forces? Unexpectedly Large Uncertainties in Molecular Datasets
Kuryla, Domantas
Berger, Fabian
Csányi, Gábor
Michaelides, Angelos
Chemical Physics
Training of general-purpose machine learning interatomic potentials (MLIPs) relies on large datasets with properties usually computed with density functional theory (DFT). A pre-requisite for accurate MLIPs is that the DFT data are well converged to minimize numerical errors. A possible symptom of errors in DFT force components is nonzero net force. Here, we consider net forces in datasets including SPICE, Transition1x, ANI-1x, ANI-1xbb, AIMNet2, QCML, and OMol25. Several of these datasets suffer from significant nonzero DFT net forces. We also quantify individual force component errors by comparison to recomputed forces using more reliable DFT settings at the same level of theory, and we find significant discrepancies in force components averaging from 1.7 meV/Å in the SPICE dataset to 33.2 meV/Å in the ANI-1x dataset. These findings underscore the importance of well converged DFT data as increasingly accurate MLIP architectures become available.
title How Accurate Are DFT Forces? Unexpectedly Large Uncertainties in Molecular Datasets
topic Chemical Physics
url https://arxiv.org/abs/2510.19774