EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Attention Mechanisms in Physics-Inspired Graph Neural Networks for the Max-Cut Problem

Temuujin Delgertsogt, Dalaijargal Purevsuren, Gantulga Gombojav

ICT Focus · 08.08.2026 · 5 (1)

DOI: 10.58873/sict.v5i1.45 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ICT Focus, online erschienen am 08.08.2026. Verfasst von Temuujin Delgertsogt und 2 weiteren Autor:innen. Die Arbeit nennt 32 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Physics-Inspired Graph Neural Networks (PI-GNNs) reformulate MAX-CUT as QUBO energy minimization, training a GNN to produce soft binary node assignments without labeled data. The baseline PI-GCN uses static, degree-normalized aggregation, while its attention-augmented counterpart PI-GAT---built on GATv2---introduces additional hyperparameters whose effects remain uncharacterized. This paper addresses that gap through controlled experiments on five Gset benchmark graphs (800--10,000 nodes). The first experiment examines how attention dropout p_attn affects solution quality and convergence stability across 25 seeds per configuration. The second sweeps depth from 1 to 20 layers, quantifying oversmoothing via Dirichlet energy, cosine similarity, and mean absolute deviation. On dense graphs (average degree 10), dropout in [0.10, 0.20] reduces inter-run variance by 30--41% while improving best-cut value; on the sparse graph G70 (average degree 2, diameter 34), any dropout is detrimental. PI-GAT delays oversmoothing onset by one to two layers versus PI-GCN, at 1.5-1.8x computational overhead per layer. These findings establish graph density and diameter as the primary structural determinants of whether dynamic attention benefits the PI-GNN framework.

Zitieren

APA

Delgertsogt, T., Purevsuren, D., & Gombojav, G. (2026). Attention Mechanisms in Physics-Inspired Graph Neural Networks for the Max-Cut Problem. ICT Focus, 5(1). https://doi.org/10.58873/sict.v5i1.45

Vancouver

Delgertsogt T, Purevsuren D, Gombojav G. Attention Mechanisms in Physics-Inspired Graph Neural Networks for the Max-Cut Problem. ICT Focus. 2026;5(1). doi:10.58873/sict.v5i1.45

BibTeX
@article{Delgertsogt2026,
  author = {Delgertsogt, Temuujin and Purevsuren, Dalaijargal and Gombojav, Gantulga},
  title = {Attention Mechanisms in Physics-Inspired Graph Neural Networks for the Max-Cut Problem},
  journal = {ICT Focus},
  year = {2026},
  volume = {5},
  number = {1},
  doi = {10.58873/sict.v5i1.45}
}
Kostenlos starten →