much like further exchanges to improve our work, but the following is our best effort within the current limits

We sincerely appreciate the reviewers for their careful reading, constructive questions and suggestions. First, we address questions appeared at least twice. We write P1, P2 for paragraph reference, and Rx for reviewers. That is, they only consider when the representations are precisely equal. To the best of our knowledge, our work is the first to incorporate continuous similarity into designing GNN.