All roads lead to rome: Integrating out-of-distribution detection into heterophilic graph learning
Published in IJCAI, 2025
Heterophilic graph neural networks (GNNs) have gained prominence for their ability to learn effective representations in graphs with diverse, attribute-aware relationships. While existing methods leverage attribute inference during message passing to improve performance, they often struggle with challenging heterophilic graphs. This is due to edge distribution shifts introduced by diverse connection patterns, which blur attribute distinctions and undermine message-passing stability. This paper introduces H$_2$OGNN, a novel framework that reframes edge attribute inference as an out-of-distribution (OOD) detection problem. H$_2$OGNN introduces a simple yet effective symbolic energy regularization approach for OOD learning, ensuring robust classification boundaries between homophilic and heterophilic edge attributes. This design significantly improves the stability and reliability of GNNs across diverse connectivity patterns. Through theoretical analysis, we show that H$_2$OGNN addresses the graph denoising problem by going beyond feature smoothing, offering deeper insights into how precise edge attribute identification boosts model performance. Extensive experiments on nine benchmark datasets demonstrate that H$_2$OGNN not only achieves state-of-the-art performance but also consistently outperforms other heterophilic GNN frameworks, particularly on datasets with high heterophily.
Recommended citation: Wang Y, Huang C, Li M, Cai T, Zheng Z, Huang X. All roads lead to rome: Integrating out-of-distribution detection into heterophilic graph learning[C]// Proceeding of International Joint Conference on Artificial Intelligence (IJCAI). 2025.