The authors present a neural network architecture for set functions, i.e. to identify a subset within a larger set. The authors provide a clear introduction to the problem in terms of convolutional operators and design CNN architectures on top of [1] through the addition of pooling operations for addressing the set function problem. This work tests this method on 3 synthetic problems as well several real world problems. The resulting networks performed competitively with baseline graph convolutional networks although they were outperformed slightly on subsets of tasks. The reviewers greatly appreciated the presentation of the work as the ideas were well motivated, the explanations were clear and the overall presentation were organized. Reviewers commented on the fact that the experiments were conducted on relatively small datasets. This reflected both the lack of good training data but also the scaling limitations of the proposed method. Indeed, as R2 suggests, it would be great for the authors to demonstrate the applicability and utility of this method on real world data as suggested in Section 2.1. The lack of experiments demonstrating a strong utility for the method on real world data is concerning and a detraction of the paper. The authors are strongly encouraged to pursue research in this direction. That said, given the clarity of presentation and the strong motivation that this is an important problem, the meta-reviewer is favorable to accepting this paper into this conference provided the authors address all of the points and remarks provided by the reviewers. This paper provides an important touchpoint on an interesting machine learning problem. [1] Markus Püschel. A discrete signal processing framework for set functions. In Proc. International Conference on Acoustics, Speech, and Signal Processing (ICASSP), 2018.