Get Started
Research questionHow should we choose graph counterfactual explainers when small edits, realism, coverage, and quality conflict?Graph counterfactual explainers seek graph edits that change a graph neural network’s prediction, but methods differ in how minimally, realistically, and broadly they produce useful explanations. Supporting both edge additions and removals does not eliminate these trade-offs.
AI
Evaluation & Benchmarks
Machine Learning
Latest papersRecent research connected to this question, newest first.A Comparative Study of Counterfactual Explainers for Graph Neural Networks Enabling Multiple Types of Graph EditThe evidence compares six state-of-the-art explainers on real-world and synthetic datasets spanning binary and multiclass graph and node classification. It uses quantitative and qualitative assessments of explanation size, coverage, realism, and quality; it does not establish a universally best explainer.research paper · Sep 4, 2026
Related questions
How can we generate realistic, minimally edited visual counterfactuals without inheriting classifier shortcuts?How can counterfactual image edits be audited for a shared latent explanation across regions?How can we find minimal road-network changes that make a user-selected route optimal?How can medical-image counterfactuals reveal a classifier’s decisions without inheriting a generator’s biases?
Home
Topics
Search
Library