arXiv Machine Learning
Understanding and Mitigating Under-Confidence in GNNs from the Final Layer
Read the original on arXiv Machine Learning →The Flow has not summarised this story yet — read it at arXiv Machine Learning.
Related stories
arXiv Machine Learning
Jul 14
When does distribution shift break graph neural networks calibration?
arXiv:2607. 10804v1 Announce Type: new Abstract: Graph neural networks (GNNs) are increasingly deployed in real-world applications where distribution shift is un-avoidable.
arXiv AI
Jun 9
The Confidence Trap: Calibration Attacks for Graph Neural Networks
arXiv:2606. 08467v1 Announce Type: cross Abstract: While confidence calibration is essential for trustworthy decision-making in safety-critical applications, the robustness of calibrated GNNs to adversarial structural perturbations remains largely unexplored.
arXiv Machine Learning
Oct 8
Neighborhood Smoothing for Calibration
arXiv:2610. 09020v1 Announce Type: new Abstract: Modern neural networks are often miscalibrated, with a tendency to overconfidence.
arXiv Machine Learning
4d ago
Diagnosing GNN Failure Modes with Layer-Wise Probe Entropy
The paper introduces a post-hoc diagnostic for graph neural networks that freezes a trained model, fits a linear probe to each layer, and tracks probe accuracy with normalized predictive entropy. By examining layer-wise probe entropy, the authors can distinguish beneficial, neutral, and harmful smoothing effects—something existing geometric measures cannot do. The study also shows that entropy degradation patterns separate two main GNN failure modes: structural mismatch under heterophily, which appears early even in shallow models, and excessive-depth oversmoothing that emerges only as propagation depth increases.
arXiv Machine Learning
Oct 9
How Much Does Message Passing Matter? A Drop-In Study of GNN Layers for Neural Network Graph Regression
The study investigates how different message‑passing (MP) layers affect graph‑level regression performance in Graph Neural Networks (GNNs). By replacing the MP component in four existing GNN regressors with ten varied configurations—including convolutional, isomorphism‑based, and attention‑based designs—the authors evaluate performance across eleven neural‑network graph datasets, measuring metrics such as rank correlation, prediction error, top‑k retrieval, latency, and memory usage. Results show that the choice of MP layer significantly influences outcomes, with the optimal layer depending on graph size, training‑set size, and regression objective; classical layers like GEN, k‑GNN, and PNA outperform attention‑based layers on small graphs, while GATv2 excels on datasets with few large graphs.
arXiv AI
Jul 7
Hyperparameter Transfer in Graph Neural Networks
arXiv:2607. 05017v1 Announce Type: cross Abstract: The performance of deep learning models crucially depends on the settings of hyperparameters like learning rate, initialization scale, and weight decay.