mlsanityData to 7 Oct 2026

Towards Robust Fidelity for Evaluating Explainability of Graph Neural Networks

arXiv 2310.01820 ↗First posted 3 Oct 2023cs.LGcited by 16

Not on the map. Why?

Abstract (arXiv)

Graph Neural Networks (GNNs) are neural models that leverage the dependency structure in graphical data via message passing among the graph nodes. GNNs have emerged as pivotal architectures in analyzing graph-structured data, and their expansive application in sensitive domains requires a comprehensive understanding of their decision-making processes -- necessitating a framework for GNN explainability. An explanation function for GNNs takes a pre-trained GNN along with a graph as input, to produce a `sufficient statistic' subgraph with respect to the graph label. A main challenge in studying GNN explainability is to provide fidelity measures that evaluate the performance of these explanation functions. This paper studies this foundational challenge, spotlighting the inherent limitations of prevailing fidelity metrics, including Fid_+, Fid_-, and Fid_Delta. Specifically, a formal, information-theoretic definition of explainability is introduced and it is shown that existing metrics often fail to align with this definition across various statistical scenarios. The reason is due to potential distribution shifts when subgraphs are removed in computing these fidelity measures. Subsequently, a robust class of fidelity measures are introduced, and it is shown analytically that they are resilient to distribution shift issues and are applicable in a wide range of scenarios. Extensive empirical analysis on both synthetic and real datasets are provided to illustrate that the proposed metrics are more coherent with gold standard metrics. The source code is available at https://trustai4s-lab.github.io/fidelity.

Key facts. Cited by 16 papers in the map (data through 7 Oct 2026). Builds on 35 of its 53 references that are in the map. First posted on arXiv 3 Oct 2023.

Builds on

35 of 53 references.

References · 35

  1. 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,282
  2. 2017Graph Attention Networkscited by 8,854
  3. 2017Inductive Representation Learning on Large Graphscited by 6,284
  4. 2018How Powerful are Graph Neural Networks?cited by 4,355
  5. 2019GNNExplainer: Generating Explanations for Graph Neural Networkscited by 735
  6. 2018A Benchmark for Interpretability Methods in Deep Neural Networkscited by 467
  7. 2020Parameterized Explainer for Graph Neural Networkcited by 423
  8. 2020Explainability in Graph Neural Networks: A Taxonomic Surveycited by 383
  9. 2022From Anecdotal Evidence to Quantitative Evaluation Methods: A Systematic Review on Evaluating Explainable AIcited by 299
  10. 2021On Explainability of Graph Neural Networks via Subgraph Explorationscited by 280
  11. 2020XGNN: Towards Model-Level Explanations of Graph Neural Networkscited by 211
  12. 2022Interpretable and Generalizable Graph Learning via Stochastic Attention Mechanismcited by 171
23 more
  1. 2019Explainability Techniques for Graph Convolutional Networkscited by 160
  2. 2021Generative Causal Explanations for Graph Neural Networkscited by 124
  3. 2022Trustworthy Graph Neural Networks: Aspects, Methods and Trendscited by 88
  4. 2021The Out-of-Distribution Problem in Explainability and Search Methods for Feature Importance Explanationscited by 86
  5. 2022Evaluating Explainability for Graph Neural Networkscited by 85
  6. 2021DIG: A Turnkey Library for Diving into Graph Deep Learning Researchcited by 76
  7. 2021Towards Self-Explainable Graph Neural Networkcited by 75
  8. 2021Robust Counterfactual Explanations on Graph Neural Networkscited by 72
  9. 2022Global Explainability of GNNs via Logic Combination of Learned Conceptscited by 64
  10. 2022GraphFramEx: Towards Systematic Evaluation of Explainability Methods for Graph Neural Networkscited by 57
  11. 2020Evaluations and Methods for Explanation through Robustness Analysiscited by 51
  12. 2022CLEAR: Generative Counterfactual Explanations on Graphscited by 48
  13. 2022GStarX: Explaining Graph Neural Networks with Structure-Aware Cooperative Gamescited by 42
  14. 2022GNNInterpreter: A Probabilistic Generative Model-Level Explanation for Graph Neural Networkscited by 34
  15. 2022Interpretable Geometric Deep Learning via Learnable Randomness Injectioncited by 27
  16. 2022A Survey of Explainable Graph Neural Networks: Taxonomy and Evaluation Metricscited by 26
  17. 2022Task-Agnostic Graph Explanationscited by 20
  18. 2022A Survey of Trustworthy Graph Learning: Reliability, Explainability, and Privacy Protectioncited by 20
  19. 2023MixupExplainer: Generalizing Explanations for Graph Neural Networks with Data Augmentationcited by 18
  20. 2021Resisting Out-of-Distribution Data Problem in Perturbation of XAIcited by 12
  21. 2023Don't be fooled: label leakage in explanation methods and the importance of their quantitative evaluationcited by 11
  22. 2023GInX-Eval: Towards In-Distribution Evaluation of Graph Neural Network Explanationscited by 9
  23. 2023Efficient GNN Explanation via Learning Removal-based Attributioncited by 6

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 9; 2025: 1; 2026: 6.

16 papers.

Sort:

  1. 2024Graph Machine Learning in the Era of Large Language Models (LLMs)cited by 33
  2. 2024F-Fidelity: A Robust Framework for Faithfulness Evaluation of Explainable AIcited by 17
  3. 2024Reconsidering Faithfulness in Regular, Self-Explainable and Domain Invariant GNNscited by 14
  4. 2024How Interpretable Are Interpretable Graph Neural Networks?cited by 13
  5. 2024Explaining Graph Neural Networks via Structure-aware Interaction Indexcited by 10
  6. 2024A Self-Explainable Heterogeneous GNN for Relational Deep Learningcited by 7
  7. 2025B-XAIC Dataset: Benchmarking Explainable AI for Graph Neural Networks Using Chemical Datacited by 4
  8. 2024Generating In-Distribution Proxy Graphs for Explaining Graph Neural Networkscited by 3
  9. 2026GNN Explanations that do not Explain and How to find Themcited by 3
  10. 2024On the Feasibility of Fidelity^- for Graph Pruningcited by 1
  11. 2024PAC Learnability under Explanation-Preserving Graph Perturbationscited by 0
  12. 2026Empirical Characterization of Rationale Stability Under Controlled Perturbations for Explainable Pattern Recognitioncited by 0
4 more
  1. 2026AIMing for Standardised Explainability Evaluation in GNNs: A Framework and Case Study on Graph Kernel Networkscited by 0
  2. 2026Learning Quantifiable Visual Explanations Without Ground-Truthcited by 0
  3. 2026A Polynomial Architecture-Attribution Co-Design Framework for Exact Aumann-Shapley Attribution in GNNscited by 0
  4. 2026WOMBAT: Whitebox Oracle for Molecular Benchmarking and Attribution Testingcited by 0