Concept等变GNN / Equivariant GNN

Equivariant Graph Neural Network

A graph neural network model that incorporates group equivariance constraints, so that it remains equivariant under input transformations such as rotation and translation.

Timeline (last 90 days)

Sep 2

E(3)-equivariant graph neural networks have demonstrated significant advantages in protein structure prediction and molecular property prediction.

Unverified50%

All Facts (1)

Unverified

E(3)-equivariant graph neural networks have demonstrated significant advantages in protein structure prediction and molecular property prediction.

50%

Source Articles

Equivariant Representation Spectral Symmetry Fingerprint: A New Perspective for Geometric Deep LearningNov 6