A Modal Logic for Explaining some Graph Neural Networks
This work addresses the interpretability of GNNs for researchers and practitioners in machine learning, but it appears incremental as it builds on existing logic and GNN frameworks.
The authors tackled the problem of explaining Graph Neural Networks (GNNs) by proposing a modal logic with counting modalities in linear inequalities, showing that formulas and GNNs can be transformed into each other, and proving that the satisfiability problem is decidable, with some variants in PSPACE.
In this paper, we propose a modal logic in which counting modalities appear in linear inequalities. We show that each formula can be transformed into an equivalent graph neural network (GNN). We also show that each GNN can be transformed into a formula. We show that the satisfiability problem is decidable. We also discuss some variants that are in PSPACE.