Generator-based Graph Generation via Heat Diffusion

TL;DR

This paper introduces a novel graph generation framework using heat diffusion and neural networks to model graph structures, unifying and extending existing diffusion-based methods.

Contribution

It adapts the Generator Matching paradigm to graph data using the graph Laplacian and heat kernel, enabling flexible and domain-informed graph generation.

Findings

Effectively captures structural properties of real and synthetic graphs

Unifies and generalizes existing diffusion-based graph generative models

Uses neural networks to learn and simulate graph diffusion processes

Abstract

Graph generative modelling has become an essential task due to the wide range of applications in chemistry, biology, social networks, and knowledge representation. In this work, we propose a novel framework for generating graphs by adapting the Generator Matching (arXiv:2410.20587) paradigm to graph-structured data. We leverage the graph Laplacian and its associated heat kernel to define a continous-time diffusion on each graph. The Laplacian serves as the infinitesimal generator of this diffusion, and its heat kernel provides a family of conditional perturbations of the initial graph. A neural network is trained to match this generator by minimising a Bregman divergence between the true generator and a learnable surrogate. Once trained, the surrogate generator is used to simulate a time-reversed diffusion process to sample new graph structures. Our framework unifies and generalises existing diffusion-based graph generative models, injecting domain-specific inductive bias via the Laplacian, while retaining the flexibility of neural approximators. Experimental studies demonstrate that our approach captures structural properties of real and synthetic graphs effectively.