Semifactual Explanations for GNN-based Classification: Formal Foundations, Complexity and Computation
A novel technique for post-hoc explainability queries in GNNs is introduced by focusing on the semifactual reasoning, and a novel learning architecture for addressing their computation is proposed.