The paper reviews three broad statistical perspectives on generalization in GNNs and highlights key results plus limitations for each.
arXiv preprint (arXiv:2603.17569) (2026)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Different Statistical Perspectives for Understanding Generalisation in Graph Neural Networks
The paper reviews three broad statistical perspectives on generalization in GNNs and highlights key results plus limitations for each.