Cross-attentive Cohesive Subgraph Embedding to Mitigate Oversquashing in GNNs

ArXi:2603.27529v1 Announce Type: cross Graph neural networks (GNNs) have achieved strong performance across various real-world domains. Nevertheless, they suffer from oversquashing, where long-range information is distorted as it is compressed through limited message-passing pathways. This bottleneck limits their ability to capture essential global context and decreases their performance, particularly in dense and heterophilic regions of graphs.