graph neural networks (GNNs) have achieved impressive results in graph
classification tasks, but they struggle to generalize effectively when faced
with out-of-distribution (OOD) data. Several approaches have been proposed to
address this problem. Among them, one solution is to diversi