Generating Directed Graphs with Dual Attention and Asymmetric Encoding

2025-06-19

AI Summary

Directed graph generation faces two major bottlenecks: the inherent difficulty of modeling directional asymmetry and the absence of a standardized evaluation benchmark. To address these, we propose Directo—the first directed graph generation framework based on discrete flow matching. It introduces an asymmetric positional encoding tailored for directed edges and a dual-path attention mechanism that jointly models bidirectional dependencies, enabling effective capture of non-symmetric structural relationships. Directo achieves state-of-the-art performance across diverse real-world and synthetic directed graphs, including DAGs, demonstrating strong generalization and competitive accuracy against task-specific models. Furthermore, we construct and open-source the first standardized benchmark suite for directed graph generation—comprising curated datasets, unified evaluation metrics, and comprehensive baselines—significantly enhancing reproducibility and enabling fair, systematic comparison in this emerging field.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Natural Language Processing: Generation
• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Directed graphs naturally model systems with asymmetric, ordered relationships, essential to applications in biology, transportation, social networks, and visual understanding. Generating such graphs enables tasks such as simulation, data augmentation and novel instance discovery; however, directed graph generation remains underexplored. We identify two key factors limiting progress in this direction: first, modeling edge directionality introduces a substantially larger dependency space, making the underlying distribution harder to learn; second, the absence of standardized benchmarks hinders rigorous evaluation. Addressing the former requires more expressive models that are sensitive to directional topologies. We propose Directo, the first generative model for directed graphs built upon the discrete flow matching framework. Our approach combines: (i) principled positional encodings tailored to asymmetric pairwise relations, (ii) a dual-attention mechanism capturing both incoming and outgoing dependencies, and (iii) a robust, discrete generative framework. To support evaluation, we introduce a benchmark suite covering synthetic and real-world datasets. It shows that our method performs strongly across diverse settings and even competes with specialized models for particular classes, such as directed acyclic graphs. Our results highlight the effectiveness and generality of our approach, establishing a solid foundation for future research in directed graph generation.

Problem

• Modeling complex dependencies in directed graph generation
• Lack of standardized benchmarks for evaluation
• Need for expressive models sensitive to directional topologies

Innovation

• Dual-attention mechanism for directional dependencies
• Asymmetric positional encodings for pairwise relations
• Discrete flow matching framework for generation