New Credit System Proposed to Improve Machine Learning Peer Review Quality

The machine learning research community faces significant challenges with its peer review system, including an overwhelming volume of submissions and widespread dissatisfaction with review quality. Current approaches, which often rely on polite guidelines, have proven insufficient to address these issues effectively. This paper argues that meaningful improvements require a more robust framework. The authors propose a novel solution centered on enforceable procedural safeguards combined with a currency-like credit system. This system, exemplified by "OpenReview Points," would allow ML practitioners to accumulate credits by contributing high-quality reviews. These earned points could then be redeemed for various benefits across major conferences, such as complimentary registration or the ability to request additional review resources.