Reproduce “Can AI Reviewers See Your Efforts to Improve Your Paper? A Large-Scale Study of Paper Revision Trajectories” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yaochen Wang, Chenguang Wang, Ming Li, Haomin Zhuang, Xinyue Zeng, Han Chen, Kwesi Adu Cobbina, Dang Nguyen and 3 more
PaperTrajectory tests whether independent AI review scores reflect improvements between paper versions and whether score gains persist under opinion cues. It separates judgments of overall quality from changes across 14 manuscript dimensions, then evaluates eight reviewers on accepted papers from five AI conferences in 2026. For GPT-6 Astra, 77.63% of 1,484 major-improvement pairs received unchanged ratings; results are limited to this dataset and tested reviewers.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.