403
Sorry!!
Error! We're sorry, but the page you were looking for doesn't exist.
OpenAI Dumps 722 AI-Written Math Papers, Warns of Errors
(MENAFN) OpenAI has released hundreds of mathematics research papers generated by an unreleased AI model, opening the findings to public scrutiny while cautioning that some may contain errors.
The collection, posted Tuesday on the code-sharing platform GitHub, comprises 722 manuscripts organized into 372 sets of related results. A set can include a headline finding, supporting arguments or alternative proofs. During testing, the model took on roughly 4,000 research problems, the company said.
OpenAI said the published results sit at varying stages of verification. For many of them, it also provided proofs written in Lean, a programming language that lets computers check whether a mathematical argument holds up logically, and it intends to add more as they become available.
The company also released 10 summaries of the model’s reasoning, along with estimates of the computing power involved. On average, each result required computing power equivalent to about three hours of ChatGPT Pro “thinking,” it said.
OpenAI said it consulted the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study on how to present and share the work.
“This is an important event for mathematics, with consequences both for mathematics and for the mathematical community that extend far beyond the individual results,” the group said, stressing its advisory role did not amount to an endorsement of the results.
“Making this work public is a first step. This release is the beginning, not the completion, of the process of human understanding and the incorporation of the work into mathematical knowledge,” it added. “The future of mathematical research cannot consist only of understanding results produced by AI labs.”
The collection, posted Tuesday on the code-sharing platform GitHub, comprises 722 manuscripts organized into 372 sets of related results. A set can include a headline finding, supporting arguments or alternative proofs. During testing, the model took on roughly 4,000 research problems, the company said.
OpenAI said the published results sit at varying stages of verification. For many of them, it also provided proofs written in Lean, a programming language that lets computers check whether a mathematical argument holds up logically, and it intends to add more as they become available.
The company also released 10 summaries of the model’s reasoning, along with estimates of the computing power involved. On average, each result required computing power equivalent to about three hours of ChatGPT Pro “thinking,” it said.
OpenAI said it consulted the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study on how to present and share the work.
“This is an important event for mathematics, with consequences both for mathematics and for the mathematical community that extend far beyond the individual results,” the group said, stressing its advisory role did not amount to an endorsement of the results.
“Making this work public is a first step. This release is the beginning, not the completion, of the process of human understanding and the incorporation of the work into mathematical knowledge,” it added. “The future of mathematical research cannot consist only of understanding results produced by AI labs.”
Legal Disclaimer:
MENAFN provides the
information “as is” without warranty of any kind. We do not accept any
responsibility or liability for the accuracy, content, images, videos,
licenses, completeness, legality, or reliability of the information
contained in this article. If you have any complaints or copyright issues
related to this article, kindly contact the provider above.
Comments
No comment