Recognition: unknown
GraphVAE: Towards Generation of Small Graphs Using Variational Autoencoders
read the original abstract
Deep learning on graphs has become a popular research topic with many applications. However, past work has concentrated on learning graph embedding tasks, which is in contrast with advances in generative models for images and text. Is it possible to transfer this progress to the domain of graphs? We propose to sidestep hurdles associated with linearization of such discrete structures by having a decoder output a probabilistic fully-connected graph of a predefined maximum size directly at once. Our method is formulated as a variational autoencoder. We evaluate on the challenging task of molecule generation.
This paper has not been read by Pith yet.
Forward citations
Cited by 2 Pith papers
-
Dependency-Aware Discrete Diffusion for Scene Graph Generation
A new discrete diffusion model for scene graph generation from text captures object-relation dependencies via hierarchical constraints and training-free conditioning, yielding better graph metrics and downstream image...
-
Equivariant Efficient Joint Discrete and Continuous MeanFlow for Molecular Graph Generation
EQUIMF is a unified equivariant framework that jointly generates discrete topologies and continuous geometries in molecular graphs via synchronized MeanFlow dynamics for efficient few-step sampling.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.