Multiresolution Equivariant Graph Variational AutoencoderDownload PDF

Published: 15 Jun 2022, Last Modified: 05 May 2023ICML-AI4Science PosterReaders: Everyone
Keywords: Graph neural networks, graph variational autoencoders, hierarchical generative models, molecule generation, supervised and unsupervised molecular representation learning
TL;DR: The first generative model that is able to generate graphs in both equivariant and multiresolution manner.
Abstract: In this paper, we propose Multiresolution Equivariant Graph Variational Autoencoders (MGVAE), the first hierarchical generative model to learn and generate graphs in a multiresolution and equivariant manner. At each resolution level, MGVAE employs higher order message passing to encode the graph while learning to partition it into mutually exclusive clusters and coarsening into a lower resolution that eventually creates a hierarchy of latent distributions. MGVAE then constructs a hierarchical generative model to variationally decode into a hierarchy of coarsened graphs. Importantly, our proposed framework is end-to-end permutation equivariant with respect to node ordering. MGVAE achieves competitive results with several generative tasks including general graph generation, molecular generation, unsupervised molecular representation learning to predict molecular properties, link prediction on citation graphs, and graph-based image generation.
Track: Original Research Track
0 Replies

Loading