Graph Neural Networks Explanation with Distill n' Explain

Explaining node predictions in GNNs using simple surrogates can improve interpretability and performance.