NeurIPS 2020

Linear Disentangled Representations and Unsupervised Action Estimation

Meta Review

In the discussion phase, reviewers expressed that, taking into account their rebuttal, authors were on a good path to have a sufficiently improved updated version for the camera ready. The AC confirms that you must include the clarifications and additions mentioned in your rebuttal (including experiments with variation with added noise/distractors and results on dSprites) in your final version. With this the work will cover sufficient interesting material, and AC and reviewers all agree that it be presented at NeurIPS.