E(n) Equivariant Normalizing Flows

Abstract

This paper introduces a generative model equivariant to Euclidean symmetries: E(n) Equivariant Normalizing Flows (E-NFs). To construct E-NFs, we take the discriminative E(n) graph neural networks and integrate them as a differential equation to obtain an invertible equivariant function: a continuous-time normalizing flow. We demonstrate that E-NFs considerably outperform baselines and existing methods from the literature on particle systems such as DW4 and LJ13, and on molecules from QM9 in terms of log-likelihood. To the best of our knowledge, this is the first flow that jointly generates molecule features and positions in 3D.

arXiv commentAccepted at Neural Information Processing Systems (NeurIPS 2021)

References in corpus (23)

• Semi-Supervised Classification with Graph Convolutional Networks
• NICE: Non-linear Independent Components Estimation
• Variational Inference with Normalizing Flows
• Activation Functions: Comparison of trends in Practice and Research for Deep Learning
• Group Equivariant Convolutional Networks
• Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds
• GraphRNN: Generating Realistic Graphs with Deep Auto-regressive Models
• Flow++: Improving Flow-Based Generative Models with Variational Dequantization and Architecture Design
• Efficient Graph Generation with Graph Recurrent Attention Networks
• Exploiting Cyclic Symmetry in Convolutional Neural Networks
• Sampling using $SU(N)$ gauge equivariant flows
• E(n) Equivariant Graph Neural Networks
• How to train your neural ODE: the world of Jacobian and kinetic regularization
• Symmetry-adapted generation of 3d point sets for the targeted discovery of molecules
• High-Dimensional Probability Estimation with Deep Density Models
• Equivariant Flows: sampling configurations for multi-body systems with symmetric energies
• SurVAE Flows: Surjections to Bridge the Gap between VAEs and Flows
• Equivariant Hamiltonian Flows
• Graph Normalizing Flows
• LieTransformer: Equivariant self-attention for Lie Groups
• Continuous-Time Flows for Efficient Inference and Density Estimation
• Learning Neural Generative Dynamics for Molecular Conformation Generation
• Neural Canonical Transformation with Symplectic Flows

Cited by in corpus (3)

• Crystal Diffusion Variational Autoencoder for Periodic Material Generation
• Robust Partial-to-Partial Point Cloud Registration in a Full Range
• Optimal Design of Experiments for Simulation-Based Inference of Mechanistic Acyclic Biological Networks