Model-Level GNN Explanations via Rule-to-Graph Readout for Logit Reconstruction

ArXi:2503.09051v2 Announce Type: replace-cross We propose a novel model-level GNN explanation framework that shifts the explanation target from class-wise rule extraction to rule-based logit reconstruction. Our method recasts the graph-level readout of a pretrained GNN as a weighted rule-level readout: grounded subgraph concepts are composed into logical rules, rule embeddings are computed directly from their symbolic structure, and active rules are passed through the frozen classifier head to reconstruct the GNN's raw multiclass logits.