Conference paper · ICML 2025
Unifews: You Need Fewer Operations for Efficient Graph Neural Networks
ICML 2025 · PMLR 267: 37587–37609Published
TL;DR
A unified, entry-wise sparsification framework that jointly prunes graph propagation and model weights across GNN layers, reducing computation while bounding approximation error.
Abstract
Graph Neural Networks (GNNs) have shown promising performance, but at the cost of resource-intensive operations on graph-scale matrices. To reduce computational overhead, previous studies attempt to sparsify the graph or network parameters, but with limited flexibility and precision boundaries. In this work, we propose Unifews, a joint sparsification technique to unify graph and weight matrix operations and enhance GNN learning efficiency. The Unifews design enables adaptive compression across GNN layers with progressively increased sparsity, and is applicable to a variety of architectures with on-the-fly simplification. Theoretically, we establish a novel framework to characterize sparsified GNN learning in view of the graph optimization process, showing that Unifews effectively approximates the learning objective with bounded error and reduced computational overhead. Extensive experiments demonstrate that Unifews achieves efficiency improvements with comparable or better accuracy, including 10-20x matrix operation reduction and up to 100x acceleration on graphs up to billion-edge scale.
Abstract from the official PMLR proceedings record. PMLR proceedings (opens in a new tab)
BibTeX
@InProceedings{pmlr-v267-liao25h,
title = {Unifews: You Need Fewer Operations for Efficient Graph Neural Networks},
author = {Liao, Ningyi and Yu, Zihao and Zeng, Ruixiao and Luo, Siqiang},
booktitle = {Proceedings of the 42nd International Conference on Machine Learning},
pages = {37587--37609},
year = {2025},
editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry},
volume = {267},
series = {Proceedings of Machine Learning Research},
month = {13--19 Jul},
publisher = {PMLR},
pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/liao25h/liao25h.pdf},
url = {https://proceedings.mlr.press/v267/liao25h.html}
}
Official PMLR citation.
Select the citation to copy it, or download the BibTeX file.