Examples for Generating Explanations of Graph Neural Networks

This directory contains examples demonstrating the use of the torch_geometric.explain package. The explain package of PyG provides a set of tools to explain the predictions of a GNN model or to explain the underlying phenomenon of a dataset.

Example	Description
gnn_explainer.py	GNNExplainer for node classification
gnn_explainer_link_pred.py	GNNExplainer for link prediction
gnn_explainer_ba_shapes.py	GNNExplainer applied on the BAShapes dataset
captum_explainer.py	Captum-based explainer for node classification
captum_explainer_hetero_link.py	Captum-based explainer for heterogenous link prediction
graphmask_explainer.py	GraphMaskExplainer for node classification