3 papers
cs.LG2026
Overcoming Shortcut Learning in Graph Neural Networks through Active Explanation Guidance
Taraneh Younesian, Steve Azzolin, Antonio Longa +3
Graph Neural Networks (GNNs) can solve prediction tasks by unintentionally exploiting shortcuts---that is, edges, nodes, and features that correlate with but are not causal for the…
cs.LG2025
Hierarchical Conformal Classification
Floris den Hengst, Inès Blin, Majid Mohammadi +2
Conformal prediction (CP) is a powerful framework for quantifying uncertainty in machine learning models, offering reliable predictions with finite-sample coverage guarantees. When…
cs.LG2025
GRAPES: Learning to Sample Graphs for Scalable Graph Neural Networks
Taraneh Younesian, Daniel Daza, Emile van Krieken +2
Graph neural networks (GNNs) learn to represent nodes by aggregating information from their neighbors. As GNNs increase in depth, their receptive field grows exponentially, leading…