← Search

Daniel Sturm

2 accepted papers

2023

Adversarial Training for Graph Neural Networks: Pitfalls, Solutions, and New Directions

NeurIPS 2023poster

Despite its success in the image domain, adversarial training did not (yet) stand out as an effective defense for Graph Neural Networks (GNNs) against graph structure perturbations. In the pursuit of fixing adversarial training (1) we show and overcome fundamental theoretical as well as practical l…

Cited by 34SourcePDFScholar