aiqbench analysis
This paper introduces SCARCE, a novel framework for label-free graph representation learning that enhances out-of-distribution generalization by using adversarial confusion to suppress spurious correlations. By theoretically linking environment confusion to feature reweighting, the method achieves state-of-the-art performance on drug discovery datasets without relying on class labels during training. This advancement addresses a critical bottleneck in deploying graph neural networks for real-world scenarios where distribution shifts are common and labeled data is scarce.
Implicit Feature Reweighting for Label-Free Graph Generalization
arXiv:2610.29571v1 Announce Type: new Abstract: Graph-level out-of-distribution (OOD) generalization is critical for deploying graph neural networks in real-world applications where distribution shifts are inevitable. Existing approaches typically require class labels during representation learning to decide which substructures are predictive, limiting their applicability. We propose SCARCE, a label-free representation-learning framework that yields distribution-shift-robust graph representations from an environment partition of the training graphs alone, whether that partition comes from dataset metadata or, when none is available, from clustering a self-supervised embedding. Our method operates in two stages: (1) a variational graph autoencoder is warm-started for graph reconstruction; and (2) an adversarial confusion mechanism forces the encoder to produce representations from which a discriminator cannot distinguish environments, while preserving graph-reconstruction fidelity. We provide theoretical analysis showing that environment confusion contracts environment-discriminative directions in the representation space, which, under regularized downstream classifiers, is equivalent to imposing a stronger directional penalty on environment-associated features. This feature-reweighting mechanism reduces reliance on spurious correlations without erasing information, offering a label-free, continuous relaxation of last-layer reweighting methods. Experiments on a controlled synthetic datasets and the DrugOOD datasets show that SCARCE achieves state-of-the-art OOD generalization among label-free methods, using no class labels during representation learning.
Abstract:Graph-level out-of-distribution (OOD) generalization is critical for deploying graph neural networks in real-world applications where distribution shifts are inevitable. Existing approaches typically require class labels during representation learning to decide which substructures are predictive, limiting their applicability. We propose SCARCE, a label-free representation-learning framework that yields distribution-shift-robust graph representations from an environment partition of the training graphs alone, whether that partition comes from dataset metadata or, when none is available, from clustering a self-supervised embedding. Our method operates in two stages: (1) a variational graph autoencoder is warm-started for graph reconstruction; and (2) an adversarial confusion mechanism forces the encoder to produce representations from which a discriminator cannot distinguish environments, while preserving graph-reconstruction fidelity. We provide theoretical analysis showing that environment confusion contracts environment-discriminative directions in the representation space, which, under regularized downstream classifiers, is equivalent to imposing a stronger directional penalty on environment-associated features. This feature-reweighting mechanism reduces reliance on spurious correlations without erasing information, offering a label-free, continuous relaxation of last-layer reweighting methods. Experiments on a controlled synthetic datasets and the DrugOOD datasets show that SCARCE achieves state-of-the-art OOD generalization among label-free methods, using no class labels during representation learning.
Machine Learning (cs.LG)
arXiv:2610.29571 [cs.LG]
(or arXiv:2610.29571v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.29571 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 286 words
Related coverage
Beyond Reasoning Gains: Mitigating General-Capability Forgetting in Large Reasoning Models
arXiv · Oct 27, 2026
Chaos Is a LADDER: Domain Generalization Beyond Invariance via Reweighting
arXiv · Jul 30, 2026
S$^3$R: Fast and Faithful RL Rollouts via Self-Sparse Speculative Decoding
arXiv · Oct 12, 2026
CAER: Causal Action Effect Reweighting for World Model Training
arXiv · Sep 1, 2026