Reproduce “Frontier Models Need Only One Text to Unmask Its Human or AI Author” 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
Atrey Desai, Deniz Bölöni-Turgut, Ty Brassington, Minjoon Choi, Jenna Russell, Rishanth Rajendhran, Bradley Emi, Hila Gonen and 4 more
The authors test whether models can attribute single texts to human authors or AI labs without web access, tools, or candidate lists. Using zero-shot prompts, they assess prose, ideas, AI-source, and author attribution. Claude Fable 5.2 names authors of 56% of 398 private reviews, but these came from six writers, limiting what the score says about peer review generally.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.