Reproduce “AAAI-26 Dual Submissions: Novel Challenges” 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
Kiri L. Wagstaff, Joydeep Biswas, Erich Merrill III, Bo An, Ida Camacho, David J. Crandall, Matthew E. Taylor
AAAI-26 organizers observed a rise in sophisticated dual submissions, including obfuscated duplicates and "alternative universe" papers, likely facilitated by generative AI tools. They implemented a multi-stage detection pipeline, leveraging abstract similarity, an LLM for full-text overlap assessment, and manual review, resulting in 141 desk-rejections for policy violations.
Key insights
- Generative AI tools are suspected to be a major factor in the increasing prevalence and sophistication of dual submissions, making them harder for human reviewers to identify.
- A multi-stage detection methodology, combining semantic similarity with LLM-based full-text analysis, can effectively triage and prioritize a large volume of submissions for human review.
- Community-wide coordination, clear policy refinements, and the development of advanced detection tools through adversarial challenges are essential to combat evolving forms of publication misconduct.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.