AI audit of 19 papers: 97.7% of comments identified real issues

littmath · x · 2026-08-27

Daniel Litt published the overall results of an automated review of his papers by Refine. Across 19 works, the AI generated 266 detailed comments: 239 were correct, 21 partially correct, and 6 incorrect, meaning 97.7% identified a real issue. Most findings were local errors, with only 2 theorem-preserving defects and 5 technical corrections to main results found. No fundamental failures were detected.

Related event: Mathematician Daniel Litt Audits His Own 19 Papers with AI: 97.7% of Comments Flag Real Issues(6 posts)→

Original post →

More from Research

Research channel →