Tag
This paper introduces the Swarm-Inspired Emergent Synchronizer (SIES), a graph-dynamical framework that learns generalizable local interaction rules for controllable collective organization, applicable to synchronization control and heterophilous graph representation learning.