Graph Neural Networks (GNNs) achieve significant performance for various learning tasks on geometric data due to the incorporation of graph structure into node representations, which renders their comprehension challenging. In this paper, we first propose a unified framework satisfied by most existing GNN explainers. Then, introduce GraphSVX, post hoc local model-agnostic explanation method spe...

We discuss a subscription game in which service providers (e.g., museums) team up in offering a limited time subscription or access pass allowing unlimited usage of their services. In this game, a natural way to allocate the subscription income among the service providers is by using the Shapley value. We show that, for the particular game considered, the Shapley value takes a very intuitive an...