OpenAI solved over 700 open math problems using an internal frontier model.
overview of all results is here:
https://github.com/openai/math/blob/main/overview.pdf
Several are fields medal worthy results. One proves the Riemann hypothesis for Re(z) >7/8, which is a substantial improvement over previous results
https://openai.com/index/sharing-ai-progress-in-mathematics/Open linkView original on lemmy.world
4 replies
I watched an online lecture recently where it was hinted that mathematicians have access to models that are not public yet and they seem impressed with their capabilities.
https://www.youtube.com/watch?v=9EuNUe-CJRM
Downvotes are disappointing; this is very clearly relevant math news, even if we don't like the news.
it will take some time for people to accept it. we are currently in the burry head in sand phase. a lot of these are prize worthy results....
Indeed. Unique games surprised us.