Wu, Y.; Bojchevski, A.; Huang, H. Adversarial Weight Perturbation Improves Generalization in Graph Neural Networks. AAAI 2023, 37, 10417-10425.