Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Raising the Bar in Graph OOD Generalization: Invariant Learning Beyond Explicit Environment Modeling

About

Out-of-distribution (OOD) generalization has emerged as a critical challenge in graph learning, as real-world graph data often exhibit diverse and shifting environments that traditional models fail to generalize across. A promising solution to address this issue is graph invariant learning (GIL), which aims to learn invariant representations by disentangling label-correlated invariant subgraphs from environment-specific subgraphs. However, existing GIL methods face two major challenges: (1) the difficulty of capturing and modeling diverse environments in graph data, and (2) the semantic cliff, where invariant subgraphs from different classes are difficult to distinguish, leading to poor class separability and increased misclassifications. To tackle these challenges, we propose a novel method termed Multi-Prototype Hyperspherical Invariant Learning (MPHIL), which introduces two key innovations: (1) hyperspherical invariant representation extraction, enabling robust and highly discriminative hyperspherical invariant feature extraction, and (2) multi-prototype hyperspherical classification, which employs class prototypes as intermediate variables to eliminate the need for explicit environment modeling in GIL and mitigate the semantic cliff issue. Derived from the theoretical framework of GIL, we introduce two novel objective functions: the invariant prototype matching loss to ensure samples are matched to the correct class prototypes, and the prototype separation loss to increase the distinction between prototypes of different classes in the hyperspherical space. Extensive experiments on 11 OOD generalization benchmark datasets demonstrate that MPHIL achieves state-of-the-art performance, significantly outperforming existing methods across graph data from various domains and with different distribution shifts.

Xu Shen, Yixin Liu, Yili Wang, Rui Miao, Yiwei Dai, Shirui Pan, Yi Chang, Xin Wang• 2025

Related benchmarks

TaskDatasetResultRank
molecule property predictionDrugOOD-IC50 (assay)
Average Accuracy72.96
40
molecule property predictionDrugOOD-EC50 (assay)
Average Accuracy78.08
40
Graph ClassificationGOOD-Motif (basis)
Accuracy76.23
31
Graph ClassificationGOOD-Motif size
Accuracy58.43
31
molecule property predictionDrugOOD-EC50 (scaffold)
Average Accuracy68.34
20
molecule property predictionDrugOOD-IC50 (scaffold)
Average Accuracy68.62
20
Graph ClassificationGOOD-HIV scaffold
Average Accuracy73.94
19
Graph ClassificationGOOD-HIV size
Accuracy66.84
19
Image ClassificationGOOD-CMNIST color
Average Accuracy41.29
17
Graph ClassificationGOOD-PCBA scaffold
Average Accuracy25.73
16
Showing 10 of 14 rows

Other info

Follow for update