Research map / Graph learning / GNN architectures
GNNBoundary: Towards Explaining Graph Neural Networks through the Lens of Decision Boundaries
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes GNNBoundary, a method that explains GNN classifiers by analyzing the decision boundaries between class pairs to reveal class-level discriminative graph patterns.
Problem
Explaining GNN decision-making, which prior work has mostly done per instance or per class, without characterizing the decision boundaries between classes.
Approach
Studies the decision boundaries of GNN classifiers, examining graphs that lie near the boundary between two classes to explain what distinguishes them, rather than explaining single predictions or class-wise discriminative patterns alone.
Closest papers by meaning
- GOAt: Explaining Graph Neural Networks via Graph Output Attribution (ICLR 2024 · similarity 0.90)
- Towards Robust Fidelity for Evaluating Explainability of Graph Neural Networks (ICLR 2024 · similarity 0.90)
- TREE-G: Decision Trees Contesting Graph Neural Networks (AAAI 2024 · similarity 0.90)
- On the Interplay between Graph Structure and Learning Algorithms in Graph Neural Networks (ICML 2025 · similarity 0.89)
- What do Graph Neural Networks learn? Insights from Tropical Geometry (NeurIPS 2024 · similarity 0.89)
- EiG-Search: Generating Edge-Induced Subgraphs for GNN Explanation in Linear Time (ICML 2024 · similarity 0.89)
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.