Explainability Techniques for Graph Convolutional Networks
This work addresses the need for interpretability in graph networks, which is incremental as it applies existing explainability methods to a specific domain.
The paper tackled the problem of explaining decisions made by Graph Convolutional Networks by evaluating gradient-based and decomposition-based techniques on a toy dataset and a chemistry task, but did not report concrete numerical results.
Graph Networks are used to make decisions in potentially complex scenarios but it is usually not obvious how or why they made them. In this work, we study the explainability of Graph Network decisions using two main classes of techniques, gradient-based and decomposition-based, on a toy dataset and a chemistry task. Our study sets the ground for future development as well as application to real-world problems.