NeurIPS 2020

Self-Supervised Graph Transformer on Large-Scale Molecular Data

Meta Review

This paper presents a few ideas for self-supervised pre-training of GNNs as well as a new GNN + transformer model, the results look solid when compared against other baselines. The reviewers have a lukewarm reaction to the paper but unanimously support an accept. I hope the authors can improve the clarity of the paper and include the additional results obtained during rebuttal into the paper in the final version.