EUVIMEDEUVIMED entdecken
Crossref

Expressivity of Graph Neural Networks: ReLU vs. Sigmoid GNNs

Sammy Khalife, Josué Tonelli-Cueto

ACM Communications in Computer Algebra · 01.12.2025 · 59 (4) · 77-80

DOI: 10.1145/3832046.3832049 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ACM Communications in Computer Algebra, online erschienen am 18.08.2026. Verfasst von Sammy Khalife und 1 weiteren Autor:innen. Die Arbeit nennt 4 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Graph Neural Networks (GNNs) form a powerful computational framework for machine learning on graphs with many applications. However, many open questions exist regarding how the chosen activation function affects expressivity, trainability and generability. In this short communication, we show that ReLU GNNs are more expressive than Sigmoid GNNs. However, we also show that Sigmoid GNNs are expressive enough so that this gap is not significant in practice, meaning that the comparison between ReLU and Sigmoid GNNs should be done in terms of trainability and generability. This is work accepted and presented at ICLR'25 [3], which is of interest to ISSAC. In this short communication for ISSAC, we present briefly the result of [3] focused on the case of ReLU vs. Sigmoid GNNs. As of today, the effect of the activation function on the properties of GNNs (expressivity, trainability and generability) is widely unknown. The work in [3] focused in the expressivity of GNNs through the perspective of logical queries following [1, 2].

Zitieren

APA

Khalife, S., & Tonelli-Cueto, J. (2025). Expressivity of Graph Neural Networks: ReLU vs. Sigmoid GNNs. ACM Communications in Computer Algebra, 59(4), 77-80. https://doi.org/10.1145/3832046.3832049

Vancouver

Khalife S, Tonelli-Cueto J. Expressivity of Graph Neural Networks: ReLU vs. Sigmoid GNNs. ACM Communications in Computer Algebra. 2025;59(4):77-80. doi:10.1145/3832046.3832049

BibTeX
@article{Khalife2025,
  author = {Khalife, Sammy and Tonelli-Cueto, Josué},
  title = {Expressivity of Graph Neural Networks: ReLU vs. Sigmoid GNNs},
  journal = {ACM Communications in Computer Algebra},
  year = {2025},
  volume = {59},
  number = {4},
  pages = {77--80},
  doi = {10.1145/3832046.3832049}
}
Kostenlos starten →