Closed alirezadizaji closed 2 years ago
Hi, in node-level explanation for GNNExplainer, we could only take nodes and edges of its K-Hop subset for optimization in order to speed up the procedure (from 17.23s to 1.85s with CPU)
Hi, in node-level explanation for GNNExplainer, we could only take nodes and edges of its K-Hop subset for optimization in order to speed up the procedure (from 17.23s to 1.85s with CPU)