OpenAI has released nearly 400 AI-generated mathematical results across more than 700 manuscripts. The collection spans diverse disciplines, including number theory, geometry, combinatorics, and mathematical physics. The company provided a table of contents and abstracts, though researchers noted that navigating the repository remains a significant challenge.
Verification of the results is currently incomplete. OpenAI stated that only 300 of the 719 manuscripts have been formalized using the Lean programming language, representing approximately 42 percent of the total. The company plans to update the repository with additional formalizations as they become available.
The release has caused widespread concern among mathematicians regarding the quality and attribution of the work. Some researchers reported that certain papers were difficult to follow or lacked proper citations. OpenAI has already retracted three papers from the initial collection.
Despite concerns, some mathematicians identified high-caliber results within the release. These include progress on the Riemann hypothesis, a special case of the Hodge conjecture, and a solution to the four-dimensional Kakeya conjecture. Experts noted that some findings could potentially warrant publication in top-tier journals.
The publication has significantly impacted the academic community, with some researchers reporting that their ongoing projects or grant proposals have been rendered obsolete. The disruption appears concentrated in specific areas, such as probability, combinatorics, and theoretical computer science.
