Sign in Subscribe
Artificial neural networks

Equivariant Neural Networks for Physics and Chemistry

Strategic Insights

— 1 min read

Researchers from the University of Chicago have developed a new class of artificial neural networks called equivariant neural nets, which can efficiently incorporate symmetries into the network structure. This innovation has significant implications for applications in physics and chemistry, where symmetries play a crucial role in understanding complex phenomena. The equivariant neural nets leverage ideas from group representation theory and can be used to model systems with inherent symmetries.

Key Takeaways:

  • The research focuses on the development of equivariant neural nets, a type of artificial neural network that can efficiently incorporate symmetries into the network structure.
  • The equivariant neural nets are based on group representation theory and can be used to model systems with inherent symmetries, such as those found in physics and chemistry.
  • The research concludes that the Clebsch-Gordan transform appears in such architectures, and can play the role of an equivariant nonlinearity.
  • Equivariant neural nets have significant implications for applications in physics and chemistry, where symmetries play a crucial role in understanding complex phenomena.
  • The research has been peer-reviewed and published in the Proceedings of the National Academy of Sciences.
  • Risi Kondor from the University of Chicago is one of the researchers leading the project.
  • The equivariant neural nets can be used to model complex systems with inherent symmetries, such as those found in molecular systems.

Statistics:

  • The research was conducted by a team of researchers from the University of Chicago.
  • The research was published in the Proceedings of the National Academy of Sciences in 2025.
  • The paper was published in volume 122, issue 41, and page number is not specified.
  • The citation for the news report is: NewsRx. New Science Findings from University of Chicago Discussed (The principles behind equivariant neural networks for physics and chemistry). Journal of Engineering. October 2025; p 2584.
  • The Corresponding Author is the University of Chicago.

Sources:

  • The principles behind equivariant neural networks for physics and chemistry. Proceedings of the National Academy of Sciences, 2025;122(41).
  • Risi Kondor, Dept. of Computer Science, University of Chicago, Chicago, IL 60605.
  • Natl Acad Sciences, 2101 Constitution Ave NW, Washington, DC 20418, USA.
  • NewsRx LLC -
  • Proceedings of the National Academy of Sciences - www.nasonline.org/publications/pnas/
  • University of Chicago -
  • National Academy of Sciences - www.nasonline.org/

Read more