EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

The principles behind equivariant neural networks for physics and chemistry

Risi Kondor iD

Proceedings of the National Academy of Sciences · 14.10.2025 · 122 (41)

DOI: 10.1073/pnas.2415656122 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the National Academy of Sciences, online erschienen am 06.10.2025. Verfasst von Risi Kondor. Beteiligte Einrichtung: Department of Computer Science, The University of Chicago, Chicago, IL 60605. Die Arbeit wurde bisher 15-mal zitiert (Crossref) und nennt 66 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

A distinguishing feature of the neural network models used in Physics and Chemistry is that they must obey basic underlying symmetries, such as symmetry to translations, rotations, and the exchange of identical particles. Over the course of the last several years, the artificial neural networks community has developed a class of networks called group-equivariant neural nets that can efficiently “bake-in” such symmetries into the structure of the network itself. Equivariant neural nets leverage ideas from group representation theory and express all variables in the generalized Fourier space corresponding to the underlying group. In this article, we review this formalism and derive the general form of operations allowable in equivariant neural networks. Specifically, we discuss why the Clebsch–Gordan transform appears in such architectures, and how it can play the role of an equivariant nonlinearity.

Schlagwörter

Stichwörter der Autor:innen

Clebsch–Gordan productequivariant neural networksneural networksnoncommutative harmonic analysis

Autor:innen und Institutionen

  • Department of Computer Science, The University of Chicago, Chicago, IL 60605.
  • Department of Statistics, Committee on Computational and Applied Mathematics, The University of Chicago, Chicago, IL 60605.

ORCID-Profile: Risi Kondor

Zitieren

APA

Kondor, R. (2025). The principles behind equivariant neural networks for physics and chemistry. Proceedings of the National Academy of Sciences, 122(41). https://doi.org/10.1073/pnas.2415656122

Vancouver

Kondor R. The principles behind equivariant neural networks for physics and chemistry. Proceedings of the National Academy of Sciences. 2025;122(41). doi:10.1073/pnas.2415656122

BibTeX
@article{Kondor2025,
  author = {Kondor, Risi},
  title = {The principles behind equivariant neural networks for physics and chemistry},
  journal = {Proceedings of the National Academy of Sciences},
  year = {2025},
  volume = {122},
  number = {41},
  doi = {10.1073/pnas.2415656122}
}
Kostenlos starten →