mlsanityData to 14 Oct 2026

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

arXiv 2503.09051 ↗First posted 12 Mar 2025cs.LG cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

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. As a result, our approach provides global explanations that remain instantiable on unseen graphs, support subgraph-level grounding, and admit rule-level contribution analysis at test-time. Experiments on three synthetic and two real-world graph classification benchmarks show that our approach faithfully reconstructs the base GNN's raw multiclass logits, achieving high probability-level fidelity across datasets. Rule-level ablations further demonstrate that the identified critical rules actively support the predicted class while suppressing non-target classes, suggesting that they act as functional units rather than merely serving as post-hoc symbolic artifacts. Compared with prior class-wise rule-based explainers, our approach achieves competitive or better prediction agreement while being up to (20×) faster, and additionally provides rule weights, test-time grounding, and logit-level contribution analysis.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 21 of its 36 references that are in the map. First posted on arXiv 12 Mar 2025.

Builds on

21 of 36 references.

References · 21

  1. 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,361
  2. 2017Inductive Representation Learning on Large Graphscited by 6,320
  3. 2018How Powerful are Graph Neural Networks?cited by 4,393
  4. 2017Modeling Relational Data with Graph Convolutional Networkscited by 1,694
  5. 2015Gated Graph Sequence Neural Networkscited by 796
  6. 2019GNNExplainer: Generating Explanations for Graph Neural Networkscited by 737
  7. 2020Parameterized Explainer for Graph Neural Networkcited by 426
  8. 2021On Explainability of Graph Neural Networks via Subgraph Explorationscited by 281
  9. 2020XGNN: Towards Model-Level Explanations of Graph Neural Networkscited by 211
  10. 2021Generative Causal Explanations for Graph Neural Networkscited by 124
  11. 2022Global Explainability of GNNs via Logic Combination of Learned Conceptscited by 65
  12. 2022GStarX: Explaining Graph Neural Networks with Structure-Aware Cooperative Gamescited by 42
9 more
  1. 2022GNNInterpreter: A Probabilistic Generative Model-Level Explanation for Graph Neural Networkscited by 34
  2. 2022Global Concept-Based Interpretability for Graph Neural Networks via Neuron Analysiscited by 32
  3. 2023DEGREE: Decomposition Based Explanation For Graph Neural Networkscited by 18
  4. 20221-WL Expressiveness Is (Almost) All You Needcited by 15
  5. 2024Explaining Graph Neural Networks via Structure-aware Interaction Indexcited by 11
  6. 2024GOAt: Explaining Graph Neural Networks via Graph Output Attributioncited by 10
  7. 2024MAGE: Model-Level Graph Neural Networks Explanations via Motif-based Graph Generationcited by 7
  8. 2022PAGE: Prototype-Based Model-Level Explanations for Graph Neural Networkscited by 5
  9. 2024EiG-Search: Generating Edge-Induced Subgraphs for GNN Explanation in Linear Timecited by 5

Built on by

Not cited yet.