The Logical Expressiveness of Temporal GNNs via Two-Dimensional Product Logics

Marco Sälzer (RPTU, University of Kaiserslautern) · Przemyslaw Walega (Queen Mary University of London, University of Łódź) · Martin Lange (University of Kassel)
expressive powerglobal tgnnsgraph neural networksgraph-and-time tgnnslogical characterisationlogical expressivenessmodal logicneural architecturespropositional temporal logicrecurrent neural networksspatial operatorsstatic gnnstemporal gnnstemporal operatorstransformerstwo-dimensional product logics

In recent years, the expressive power of various neural architectures