Geometric and Physical Quantities Improve E(3) Equivariant Message Passing
arXiv:2110.02905
Abstract
Including covariant information, such as position, force, velocity or spin is important in many tasks in computational physics and chemistry. We introduce Steerable E(3) Equivariant Graph Neural Networks (SEGNNs) that generalise equivariant graph networks, such that node and edge attributes are not restricted to invariant scalars, but can contain covariant information, such as vectors or tensors. This model, composed of steerable MLPs, is able to incorporate geometric and physical information in both the message and update functions. Through the definition of steerable node attributes, the MLPs provide a new class of activation functions for general use with steerable feature fields. We discuss ours and related work through the lens of equivariant non-linear convolutions, which further allows us to pin-point the successful components of SEGNNs: non-linear message aggregation improves upon classic linear (steerable) point convolutions; steerable messages improve upon recent equivariant graph networks that send invariant messages. We demonstrate the effectiveness of our method on several tasks in computational physics and chemistry and provide extensive ablation studies.
arXiv commentPublished at ICLR 2022 (Spotlight paper), Github: https://github.com/RobDHess/Steerable-E3-GNN
References in corpus (21)
- Deep Learning in Neural Networks: An Overview
- Decoupled Weight Decay Regularization
- Neural Message Passing for Quantum Chemistry
- Instance Normalization: The Missing Ingredient for Fast Stylization
- Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties
- The Open Catalyst 2020 (OC20) Dataset and Community Challenges
- Searching for Activation Functions
- Group Equivariant Convolutional Networks
- Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds
- Equivariant message passing for the prediction of tensorial properties and molecular spectra
- SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks
- On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups
- 3D Steerable CNNs: Learning Rotationally Equivariant Features in Volumetric Data
- Learning from Protein Structure with Geometric Vector Perceptrons
- E(n) Equivariant Graph Neural Networks
- 3D G-CNNs for Pulmonary Nodule Detection
- Relevance of Rotationally Equivariant Convolutions for Predicting Molecular Properties
- Equivariant Flows: sampling configurations for multi-body systems with symmetric energies
- Spherical Message Passing for 3D Graph Networks
- A Practical Method for Constructing Equivariant Multilayer Perceptrons for Arbitrary Matrix Groups
- An Introduction to Electrocatalyst Design using Machine Learning for Renewable Energy Storage