Adversarial Weight Perturbation Improves Generalization in Graph Neural Networks

Open in new window