ICLR 2027 sets 20-paper cap and limits first-time authors

ICLR 2027 will cap every author at 20 submissions and limit researchers with no prior paper at a major AI, ML, CV, robotics, or NLP conference to a single submission unless a co-author qualifies as a reciprocal reviewer, the program chairs announced on the ICLR blog. The volume cap is close to symbolic: "Fewer than 0.2% of authors submitted more than 20 papers last year."

The newcomer rule targets a larger cohort. About 20% of ICLR 2026 submissions had no reciprocal reviewer among the author list, and those papers "were accepted at about half the rate" of the rest; a quarter were desk-rejected. Inside that group, roughly 15% are sole-author papers and about 40% come from authors filing multiple such submissions.

The chairs frame the package as three aims: "(1) providing opportunities for 'newcomers'....(2) incentivize high-quality work, and (3) focusing precious reviewer attention on work that is likely to make an impact." They also reaffirm that all submissions become de-anonymized after review, accepted or not, arguing that "if authors are going to ask someone to put time into reviewing their paper, they should be willing to publicly associate themselves with that paper regardless of the outcome."

The post is unusually candid about who loses. "These mechanisms aren't Pareto improvements over the old reviewing system," the chairs write, naming "established interdisciplinary researchers that have published outside the reciprocal reviewing list" and junior researchers "with a backlog of high-quality papers that are being resubmitted due to low-quality reviews in previous cycles" as groups likely to feel the pinch. They add that "community members have strong feelings (both positive and negative!) about this new policy."