NeurIPS 2020

Building powerful and equivariant graph neural networks with structural message-passing


Meta Review

All reviewers appreciated the proposed new type of message passing graph neural network that aims to use node identifiers in an equivariant manner, i.e. guaranteeing that any permutation of the node ids would result in a correspondingly permuted representation of the graph. They also note the increase in computation. Overall, very relevant work for the community.