Self-inconsistency in SI-GNN explanations arises from re-explanation context perturbation; Self-Denoising post-processing calibrates explanations with one forward pass and improves quality.
Parameterized explainer for graph neural network.NeurIPS
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Why Self-Inconsistency Arises in GNN Explanations and How to Exploit It
Self-inconsistency in SI-GNN explanations arises from re-explanation context perturbation; Self-Denoising post-processing calibrates explanations with one forward pass and improves quality.