EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Regularity priors for the linear atomic cluster expansion

James P Darby, Joe D Morrow, Albert P Bartók, Volker L Deringer, Gábor Csányi, Christoph Ortner

Machine Learning: Science and Technology · 01.04.2026 · 7 (2) · 025063

DOI: 10.1088/2632-2153/ae55fa ↗

Kurz gefasst

Fachartikel in der Zeitschrift Machine Learning: Science and Technology, online erschienen am 17.04.2026. Verfasst von James P Darby und 5 weiteren Autor:innen. Die Arbeit nennt 64 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Natural Sciences and Engineering Research Council of Canada, EPSRC Centre for Doctoral Training, CASTEP-USER.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

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 atomic cluster expansion (ACE) models and explore to what extent this approach improves the quality of the fitted models. our main focus is an over-regularization that replicates the Gaussian broadening used in smooth overlap of atomic positions 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.

Förderung

  • Natural Sciences and Engineering Research Council of Canada
  • EPSRC Centre for Doctoral Training
  • CASTEP-USER
  • UK Research and Innovation
  • EPSRC

Zitieren

APA

Darby, J. P., Morrow, J. D., Bartók, A. P., Deringer, V. L., Csányi, G., & Ortner, C. (2026). Regularity priors for the linear atomic cluster expansion. Machine Learning: Science and Technology, 7(2), 025063. https://doi.org/10.1088/2632-2153/ae55fa

Vancouver

Darby JP, Morrow JD, Bartók AP, Deringer VL, Csányi G, Ortner C. Regularity priors for the linear atomic cluster expansion. Machine Learning: Science and Technology. 2026;7(2):025063. doi:10.1088/2632-2153/ae55fa

BibTeX
@article{Darby2026,
  author = {Darby, James P and Morrow, Joe D and Bartók, Albert P and Deringer, Volker L and Csányi, Gábor and Ortner, Christoph},
  title = {Regularity priors for the linear atomic cluster expansion},
  journal = {Machine Learning: Science and Technology},
  year = {2026},
  volume = {7},
  number = {2},
  pages = {025063},
  doi = {10.1088/2632-2153/ae55fa}
}
Kostenlos starten →