DropGNN: Random Dropouts Increase the Expressiveness of Graph Neural Networks

Then, we combine the results of these runs to obtain the final result. We prove that DropGNNs can distinguish various graph neighborhoods that cannot be separated by message passing GNNs.