a03fec24df877cc65c037673397ad5c0-AuthorFeedback.pdf

We are grateful to the reviewers for their substantive and constructive feedback. "brings together a few different works in an interesting way" to "paves the way for these models to keep making important contributions to our understanding of To be explicit, we used the same hyper-parameters across all experiments. Figure 1: New analysis of redundancy reduction showing similar decreases in redundancy for all learned dictionaries. We agree that the material in lines 186-191 and line 154 is less clear than it should be.