Fast Graph Generation via Autoregressive Noisy Filtration Modeling

TL;DR

This paper introduces ANFM, a novel autoregressive graph generation framework that uses topological filtration and noise strategies to produce high-quality graphs with significantly faster inference than existing diffusion models.

Contribution

ANFM combines topological filtration with noise augmentation and reinforcement learning to enable fast, high-quality, non-monotonic graph generation with error correction capabilities.

Findings

ANFM matches state-of-the-art diffusion models in quality.

ANFM achieves over 100 times faster inference.

ANFM effectively models both edge addition and deletion.

Abstract

Existing graph generative models often face a critical trade-off between sample quality and generation speed. We introduce Autoregressive Noisy Filtration Modeling (ANFM), a flexible autoregressive framework that addresses both challenges. ANFM leverages filtration, a concept from topological data analysis, to transform graphs into short sequences of subgraphs. We identify exposure bias as a potential hurdle in autoregressive graph generation and propose noise augmentation and reinforcement learning as effective mitigation strategies, which allow ANFM to learn both edge addition and deletion operations. This unique capability enables ANFM to correct errors during generation by modeling non-monotonic graph sequences. Our results show that ANFM matches state-of-the-art diffusion models in quality while offering over 100 times faster inference, making it a promising approach for high-throughput graph generation. The source code is publicly available at https://github.com/BorgwardtLab/anfm .