The reviewers are mostly satisfied by the author response, and agree about acceptance. There is still a remaining concern that additional baselines/experiments would be desirable, but given the theoretical contributions and the layer's differentiability guarantees, the contribution is already considered sufficient to warrant acceptance.