Beyond permutation equivariance in graph networks
arXiv:2103.14066
Abstract
In this draft paper, we introduce a novel architecture for graph networks which is equivariant to the Euclidean group in -dimensions. The model is designed to work with graph networks in their general form and can be shown to include particular variants as special cases. Thanks to its equivariance properties, we expect the proposed model to be more data efficient with respect to classical graph architectures and also intrinsically equipped with a better inductive bias. We defer investigating this matter to future work.
References in corpus (9)
- Neural Message Passing for Quantum Chemistry
- Group Equivariant Convolutional Networks
- Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds
- SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks
- E(n) Equivariant Graph Neural Networks
- Lorentz Group Equivariant Neural Network for Particle Physics
- Physical Symmetries Embedded in Neural Networks
- Causal Discovery in Physical Systems from Videos
- Autobahn: Automorphism-based Graph Neural Nets