NeurIPS 2020

Learning sparse codes from compressed representations with biologically plausible local wiring constraints

Meta Review

Three knowledgeable referees support acceptance for the contributions, notably for proposing a theoretical neuroscience model of randomized projections for communication between cortical areas, and I also recommend acceptance. The reviewers were happy with the rebuttal and agreed that the paper was well-backed by theoretical results and that the empirical validation was good with a balance of qualitative and quantitative results. The authors have thoughtfully engaged with the reviews, and I believe the paper will be stronger after revision.