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:2505.11335v5 Announce Type: replace Abstract: Graph Neural Networks (GNNs) have demonstrated remarkable effectiveness on graph-based tasks. However, their predictive confidence is often miscali...
arXiv:2607. 26404v1 Announce Type: new Abstract: Graph Neural Networks (GNN) facilitate effective prediction on graph data such as molecules, media networks and neural network blueprints.
arXiv:2607. 21607v1 Announce Type: cross Abstract: Graph Neural Networks propagate information through local message passing, but the graph topologies themselves can silently prevent any amount of training from solving long-range tasks.
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: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:2608. 09596v1 Announce Type: cross Abstract: Graph Neural Networks (GNNs) suffer from two fundamental limitations: over-smoothing, where node representations become indistinguishable with depth, and over-squashing, where long-range information is compressed through limited message-passing channels.
arXiv:2610.18113v1 Announce Type: cross Abstract: Graph Neural Networks (GNNs) commonly assume that node features are fully observed, despite missing features being pervasive in real-world applicatio...
arXiv:2602. 09258v2 Announce Type: replace Abstract: Deployed graph neural networks (GNNs) are frozen at deployment yet must fit clean data, generalize under distribution shifts, and remain stable to perturbations.
arXiv:2605. 21247v3 Announce Type: replace Abstract: Graph Neural Networks (GNNs) have emerged as a cornerstone of deep learning, with most existing methods rooted in graph signal processing and diffusion equations to model message passing.
arXiv:2606. 06772v1 Announce Type: cross Abstract: Understanding the generalization performance of over-parameterized neural networks has become a central topic in deep learning theory.
Scaffold is a new unsupervised graph sparsification framework for graph neural networks that uses support graph theory preconditioners to jointly control dilation and congestion, thereby preserving short communication paths while avoiding bottlenecks. It achieves superior aggregate ranking across 19 homophilic and heterophilic benchmarks, recovering or closely approaching full‑graph GNN performance with only 10%–50% of the original edges. The method reduces memory usage to less than half and cuts end‑to‑end training time, including sparsification overhead.
arXiv:2606. 08067v1 Announce Type: new Abstract: Graph neural networks (GNNs) are widely deployed on relational data, yet they can leak sensitive or proprietary information about the training graph adjacency, e.