Permutation Equivariant Generative Adversarial Networks for Graphs

TL;DR

This paper introduces 3G-GAN, a graph generative model using permutation equivariant functions within a GAN framework, aiming to address the ordering invariance problem in graph generation.

Contribution

The paper proposes a novel 3-stage GAN model called 3G-GAN that incorporates permutation equivariant functions for improved graph generation.

Findings

Initial experiments show promising results.

Highlights challenges in implementing equivariant functions.

Discusses future directions for model development.

Abstract

One of the most discussed issues in graph generative modeling is the ordering of the representation. One solution consists of using equivariant generative functions, which ensure the ordering invariance. After having discussed some properties of such functions, we propose 3G-GAN, a 3-stages model relying on GANs and equivariant functions. The model is still under development. However, we present some encouraging exploratory experiments and discuss the issues still to be addressed.