Skip to content

How Powerful are Graph Neural Networks?

← Back to topic

Authors: Keyulu Xu, Weihua Hu, Jure Leskovek, Stefanie Jegelka
Year: 2018
Journal: ICLR
DOI: 10.48550/arXiv.1810.00826
Publisher: https://arxiv.org/abs/1810.00826

Keywords: gin, gnn

Abstract

We analyze the expressive power of graph neural networks and show that they are at most as powerful as the Weisfeiler-Lehman test.

Cite this paper

bibtex
@misc{gin2018,
  title  = {How Powerful are Graph Neural Networks?},
  author = {Keyulu Xu, Weihua Hu, Jure Leskovek, Stefanie Jegelka},
  year   = {2018},
  journal = {ICLR},
  doi    = {10.48550/arXiv.1810.00826},
  url    = {https://doi.org/10.48550/arXiv.1810.00826},
}

Source files

Released under the MIT License.