Research map / Graph learning / GNN architectures
Robust Counterfactual Explanations on Graph Neural Networks
Mohit Bajaj, Lingyang Chu, Zi Yu Xue, Jian Pei, Lanjun Wang, Peter Cho-Ho Lam, Yong Zhang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add decision-boundary (LDB) loss instead of direct confidence loss (ablation) | RCExplainer · Mutagenicity, NCI1, BA-2motifs | fidelity | Improved |
| add decision-boundary (LDB) loss instead of direct confidence loss (ablation) | RCExplainer · BA-2motifs, NCI1 | noise robustness AUC | Improved |
| increase sampled LDBs from 10 to 50 (ablation) | RCExplainer · BA-Community | fidelity | Improved |
| lower lambda to 0.1 (more weight on counterfactual loss) instead of 0.9 (ablation) | RCExplainer · BA-2motifs | fidelity | Improved |
| replace per-graph explanation optimisation with decision-region-based explanation network | RCExplainer vs GNNExplainer · BA-Shapes | explanation AUC: 0.925 -> 0.998 | Improved |
| replace PGM-Explainer with RCExplainer | RCExplainer · BA-Community | node accuracy: 0.926 -> 0.916 | Hurt |
| use RCExplainer instead of PGExplainer | RCExplainer · Mutagenicity | noise robustness AUC | No clear effect |
| use trained explanation network without retraining | RCExplainer · Mutagenicity | explanation time (s): 77.8s -> 0.01s | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- PGM-Explainer — PGM-Explainer: Probabilistic Graphical Model Explanations for Graph Neural Networks (NeurIPS 2020) · better on 4 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- COMRECGC: Global Graph Counterfactual Explainer through Common Recourse (ICML 2025 · similarity 0.91)
- Towards Robust Fidelity for Evaluating Explainability of Graph Neural Networks (ICLR 2024 · similarity 0.90)
- Robust Stochastic Graph Generator for Counterfactual Explanations (AAAI 2024 · similarity 0.89)
- GNNX-BENCH: Unravelling the Utility of Perturbation-based GNN Explainers through In-depth Benchmarking (ICLR 2024 · similarity 0.88)
- Reconsidering Faithfulness in Regular, Self-Explainable and Domain Invariant GNNs (ICLR 2025 · similarity 0.88)
- Provably Robust Explainable Graph Neural Networks against Graph Perturbation Attacks (ICLR 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
