GATE: How to Keep Out Intrusive Neighbors

Nimrah Mustafa, Rebekka Burkholz
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:36990-37015, 2024.

Abstract

Graph Attention Networks (GATs) are designed to provide flexible neighborhood aggregation that assigns weights to neighbors according to their importance. In practice, however, GATs are often unable to switch off task-irrelevant neighborhood aggregation, as we show experimentally and analytically. To address this challenge, we propose GATE, a GAT extension that holds three major advantages: i) It alleviates over-smoothing by addressing its root cause of unnecessary neighborhood aggregation. ii) Similarly to perceptrons, it benefits from higher depth as it can still utilize additional layers for (non-)linear feature transformations in case of (nearly) switched-off neighborhood aggregation. iii) By down-weighting connections to unrelated neighbors, it often outperforms GATs on real-world heterophilic datasets. To further validate our claims, we construct a synthetic test bed to analyze a model’s ability to utilize the appropriate amount of neighborhood aggregation, which could be of independent interest.

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-mustafa24a, title = {{GATE}: How to Keep Out Intrusive Neighbors}, author = {Mustafa, Nimrah and Burkholz, Rebekka}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {36990--37015}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/mustafa24a/mustafa24a.pdf}, url = {https://proceedings.mlr.press/v235/mustafa24a.html}, abstract = {Graph Attention Networks (GATs) are designed to provide flexible neighborhood aggregation that assigns weights to neighbors according to their importance. In practice, however, GATs are often unable to switch off task-irrelevant neighborhood aggregation, as we show experimentally and analytically. To address this challenge, we propose GATE, a GAT extension that holds three major advantages: i) It alleviates over-smoothing by addressing its root cause of unnecessary neighborhood aggregation. ii) Similarly to perceptrons, it benefits from higher depth as it can still utilize additional layers for (non-)linear feature transformations in case of (nearly) switched-off neighborhood aggregation. iii) By down-weighting connections to unrelated neighbors, it often outperforms GATs on real-world heterophilic datasets. To further validate our claims, we construct a synthetic test bed to analyze a model’s ability to utilize the appropriate amount of neighborhood aggregation, which could be of independent interest.} }
Endnote
%0 Conference Paper %T GATE: How to Keep Out Intrusive Neighbors %A Nimrah Mustafa %A Rebekka Burkholz %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-mustafa24a %I PMLR %P 36990--37015 %U https://proceedings.mlr.press/v235/mustafa24a.html %V 235 %X Graph Attention Networks (GATs) are designed to provide flexible neighborhood aggregation that assigns weights to neighbors according to their importance. In practice, however, GATs are often unable to switch off task-irrelevant neighborhood aggregation, as we show experimentally and analytically. To address this challenge, we propose GATE, a GAT extension that holds three major advantages: i) It alleviates over-smoothing by addressing its root cause of unnecessary neighborhood aggregation. ii) Similarly to perceptrons, it benefits from higher depth as it can still utilize additional layers for (non-)linear feature transformations in case of (nearly) switched-off neighborhood aggregation. iii) By down-weighting connections to unrelated neighbors, it often outperforms GATs on real-world heterophilic datasets. To further validate our claims, we construct a synthetic test bed to analyze a model’s ability to utilize the appropriate amount of neighborhood aggregation, which could be of independent interest.
APA
Mustafa, N. & Burkholz, R.. (2024). GATE: How to Keep Out Intrusive Neighbors. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:36990-37015 Available from https://proceedings.mlr.press/v235/mustafa24a.html.

Related Material