NeurIPS 2020

A polynomial-time algorithm for learning nonparametric causal graphs

Meta Review

After discussions, the majority of reviewers think that the contributions are enough to learn causal graphs, even though some points should be made clearer in the final version (including the fact that causality may come with a stronger meaning than it is used and details on the assumptions). The response from authors has been important, but has not been appreciate by all reviewers - I suggest a better approach in the future. Moreover, I do not think that the optional field to send comments to the AC is meant to be used to write yet another rebuttal. These points could have jeopardized this submission.