ClassroomThe classroom
Back to the lesson libraryMECHANISM · 14 MIN
M34.2 CONNECT THE MECHANISM

Build symmetry into predictions about geometry

Turn a molecule 90 degrees and its energy shouldn't budge, but its forces should turn with it. See how to check that by hand and why models that know it need less data.

LESSON OVERVIEW14 min lesson

Lesson overview

Turn a molecule 90 degrees and its energy shouldn't budge, but its forces should turn with it. See how to check that by hand and why models that know it need less data.

What you’ll explore

  • Geometric learning uses invariance or equivariance under relevant transformations; match scalar, vector, and structural outputs to the symmetries the scientific task actually possesses.
GO TO THE SOURCE

Original explanations, connected to the research.

SchNet: A continuous-filter convolutional neural network for modeling quantum interactions (Schütt et al., 2017)E(n) Equivariant Graph Neural Networks (Satorras, Hoogeboom & Welling, 2021)E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials (NequIP; Batzner et al., Nature Communications 2022)Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges (Bronstein et al., 2021)
Suggest a correction

A precise note can make an explanation better.

Choose the scene and describe what needs attention. Download a feedback file to share through a channel you already use. This page does not send feedback or connect you with a reviewer.

The file includes this note, the scene title, and lesson metadata. Your saved progress and quiz responses are excluded. Download before leaving or reloading to keep your note.