[seqfan] Re: Serious LLM experimental math results

[Fred Lunnon]

  Reminds me to mention a story from a year ago, with articles in Nature &
Quanta
magazines, which I don't recall noticing here or elsewhere at the time.
    "This is a game changer! (AlphaTensor by DeepMind explained)"
    https://www.youtube.com/watch?v=3N3Bl5AA5QU
Not involving LLM, rather just old-fashioned chess- & go-playing AI trying
its hand
at optimising matrix product.

 WFL


On Fri, Dec 15, 2023 at 1:31 AM Sean A. Irvine <sairvin at gmail.com> wrote:

> Hi,
>
> At the risk of stirring up this topic yet again, here is a nice article (to
> be published in Nature) showing how a LLM was able to improve known results
> for bin packing and cap set problems.
>
> https://www.nature.com/articles/s41586-023-06924-6
>
> Sean.
>
> --
> Seqfan Mailing list - http://list.seqfan.eu/
>