Hierarchical Discrete Flow Matching for Graph Generation

2026-03-31

AI Summary

Existing denoising-based graph generation models face computational bottlenecks due to quadratic complexity in the number of nodes and the requirement for a large number of function evaluations. This work proposes a novel hierarchical graph generation framework that, for the first time, integrates hierarchical structure with discrete flow matching. By substantially reducing both the number of node pairs requiring evaluation and the number of denoising iterations, the method achieves highly efficient graph synthesis. It maintains high-quality graph structure generation while dramatically accelerating the generation process, thereby effectively overcoming the computational efficiency limitations inherent in conventional approaches.

Abstract

Denoising-based models, including diffusion and flow matching, have led to substantial advances in graph generation. Despite this progress, such models remain constrained by two fundamental limitations: a computational cost that scales quadratically with the number of nodes and a large number of function evaluations required during generation. In this work, we introduce a novel hierarchical generative framework that reduces the number of node pairs that must be evaluated and adopts discrete flow matching to significantly decrease the number of denoising iterations. We empirically demonstrate that our approach more effectively captures graph distributions while substantially reducing generation time.

Problem

• graph generation
• computational cost
• denoising iterations
• flow matching
• scalability

Innovation

• hierarchical generation
• discrete flow matching
• graph generation
• denoising-based models
• computational efficiency