Understanding convolution on graphs via energies
About
Graph Neural Networks (GNNs) typically operate by message-passing, where the state of a node is updated based on the information received from its neighbours. Most message-passing models act as graph convolutions, where features are mixed by a shared, linear transformation before being propagated over the edges. On node-classification tasks, graph convolutions have been shown to suffer from two limitations: poor performance on heterophilic graphs, and over-smoothing. It is common belief that both phenomena occur because such models behave as low-pass filters, meaning that the Dirichlet energy of the features decreases along the layers incurring a smoothing effect that ultimately makes features no longer distinguishable. In this work, we rigorously prove that simple graph-convolutional models can actually enhance high frequencies and even lead to an asymptotic behaviour we refer to as over-sharpening, opposite to over-smoothing. We do so by showing that linear graph convolutions with symmetric weights minimize a multi-particle energy that generalizes the Dirichlet energy; in this setting, the weight matrices induce edge-wise attraction (repulsion) through their positive (negative) eigenvalues, thereby controlling whether the features are being smoothed or sharpened. We also extend the analysis to non-linear GNNs, and demonstrate that some existing time-continuous GNNs are instead always dominated by the low frequencies. Finally, we validate our theoretical findings through ablations and real-world experiments.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Node Classification | Cora | Accuracy87.61 | 885 | |
| Node Classification | Citeseer | Accuracy76.92 | 804 | |
| Node Classification | Citeseer (test) | Accuracy0.773 | 729 | |
| Node Classification | Cora (test) | Mean Accuracy88.01 | 687 | |
| Node Classification | Chameleon | Accuracy71.38 | 549 | |
| Node Classification | PubMed (test) | Accuracy90.04 | 500 | |
| Node Classification | Squirrel | Accuracy59.01 | 500 | |
| Node Classification | Cornell | Accuracy84.05 | 426 | |
| Node Classification | Texas | Accuracy88.38 | 410 | |
| Node Classification | Wisconsin | Accuracy88.83 | 410 |