Wilfried N. Gansterer — accepted AI-conference papers

2025

On the Relationship Between Robustness and Expressivity of Graph Neural Networks

Lorenz Kummer, Wilfried N. Gansterer, Nils Morten Kriege

AISTATS 2025poster

We investigate the vulnerability of Graph Neural Networks (GNNs) to bit-flip attacks (BFAs) by introducing an analytical framework to study the influence of architectural features, graph properties, and their interaction. The expressivity of GNNs refers to their ability to distinguish non-isomorphic…

Cited by 2 Source ↗Scholar ↗

2025

Weisfeiler and Leman Go Gambling: Why Expressive Lottery Tickets Win

Lorenz Kummer, Samir Moustafa, Anatol Ehrlich, Franka Bause, Nikolaus Suess, Wilfried N. Gansterer +1

ICML 2025poster

The lottery ticket hypothesis (LTH) is well-studied for convolutional neural networks but has been validated only empirically for graph neural networks (GNNs), for which theoretical findings are largely lacking. In this paper, we identify the expressivity of sparse subnetworks, i.e. their ability to…

Cited by 0 Source ↗PDF ↗Scholar ↗