OpenAI drops another batch of mathematical breakthroughs

OpenAI has released a collection of mathematical manuscripts and supporting proof artifacts produced by an internal frontier model. The release includes 722 manuscripts, covering 372 result families that group related papers. These results include solutions to hundreds of open questions, according to AGMAI, an independent advisory group of elite mathematicians.

The results were obtained using an unreleased internal OpenAI model, with each result using on average three hours of ChatGPT Pro thinking compute. The model was posed approximately 4,000 problems during the evaluation. The manuscripts are organized into families, which group related papers, including principal results, companion arguments, consequences, or alternative proofs.

The release of these results has raised questions about research ethics and academic conduct, while impressing and unsettling parts of the mathematical community. According to The Verge, an independent advisory group was formed to communicate the results responsibly. OpenAI plans to continue updating the repository with formalizations and exploring community-hosted repositories for these materials.

Brief written by urgent.news from The Verge, Techmeme, Hacker News — 3 reports on this story. Machine-written — may contain errors; check the original before relying on it.