Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

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.

Francesco Di Giovanni, James Rowbottom, Benjamin P. Chamberlain, Thomas Markovich, Michael M. Bronstein• 2022

Related benchmarks

TaskDatasetResultRank
Node ClassificationCora
Accuracy87.61
1215
Node ClassificationCiteseer
Accuracy76.92
931
Node ClassificationCora (test)
Mean Accuracy88.01
861
Node ClassificationCiteseer (test)
Accuracy0.773
824
Node ClassificationChameleon
Accuracy71.38
640
Node ClassificationWisconsin
Accuracy88.83
627
Node ClassificationTexas
Accuracy88.38
616
Node ClassificationSquirrel
Accuracy59.01
591
Node ClassificationCornell
Accuracy84.05
582
Node ClassificationPubMed (test)
Accuracy90.04
546
Showing 10 of 31 rows

Other info

Follow for update