A graph neural network framework that maps graphs with arbitrary feature spaces to a shared fixed-dimension space via a structural interaction graph, enabling zero-shot transfer across graphs and recovering knowledge-graph foundation models as a special case.
International Conference on Learning Representations , year =
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Neural Message Passing on Structural Interaction Graphs for Fully-Inductive Graph Neural Networks
A graph neural network framework that maps graphs with arbitrary feature spaces to a shared fixed-dimension space via a structural interaction graph, enabling zero-shot transfer across graphs and recovering knowledge-graph foundation models as a special case.