I proved (using AI) another conjecture of mine from 2018!
This one is short enough to state in an x post so I’ll try to do that here:
Say you have an n*n matrix with nonnegative entries such that each of its row sums equals 1.
Such a matrix is called stochastic.
Assume also that its rank is k.
The conjecture (now theorem) finds the maximal value that the product of the elements of its main diagonal can achieve.
It also classify all the matrices that actually achieve that maximum.
1
118
I published a paper I’m most proud of so far.
A new partition theorem for vectors in normed spaces.
Corollaries include sharp permanent bounds for real and complex matrices of bounded rank, bounds on entropy, mutual information and more.
yair-lavi.com/publications/v…
1
66
They are making themselves irrelevant. The solution is quite easy but they will not go for it for political reasons inside “the community” - (the anti AI vibe)
Have two independent frontier models review the submission, for correctness, readability, priority etc..
arXiv has updated our policy on rate limiting for all submitters.
This update was made to fairly distribute moderator time & support the arXiv community of staff, volunteers, readers & authors.
Please read our announcement to learn more: blog.arxiv.org/2026/10/01/up…
101
Seriously?
Why is arxiv trying to slow down science?
To be clear - not one mistake was found in any of my publications.
In fact I got positive reviews on both the math and the expositions.
And now they decided to limit me to one submission a month!
However, I am able, using AI, to get several big results each week.
Really annoying!
Luckily I have zenodo and my own website
yair-lavi.com
Yair Lavi retweeted
New on the Science Blog: Yes, Claude can do Nine Loops.
Theoretical physicists predict how particles behave using formulas called scattering amplitudes. These are notoriously hard to compute, so researchers work with layers of increasingly fine corrections called “loops”—each added loop makes the answer more precise but takes exponentially more computation. Most calculations stop at two or three loops. Eight loops was the previous record in a simplified model physicists use as a testing ground (planar N=4 super-Yang-Mills), set by SLAC's Lance Dixon and collaborators.
Last month, physicist and science writer @4gravitons issued a challenge: could an AI push past eight loops in this model, using only the compute budget an academic could reasonably access?
Given a single prompt describing the nine-loop problem, Claude ran largely unsupervised for days in Claude Science and solved it using methods developed by Dixon and his colleagues, at a total cost of a few thousand dollars. Dixon independently verified the result, and von Hippel wrote about the experience for our blog.
Read more: anthropic.com/research/yes-c…
I just submitted a proof of the Marcus-Minc 1967 conjecture.
This is now the 5th conjecture/problem from the famous Minc’s list of open conjectures and problems on the permanent function I was able to prove using AI.
(Actually the 6th but someone beat me to publication..)
1
116
Yair Lavi retweeted
Claude has discovered a previously unknown enzyme system hidden in the DNA of bacteriophages. Beside the enzyme’s gene sits a long array of repeating DNA—a structure that looks somewhat similar to CRISPR.
We don’t yet understand what this system does, but only a handful of known systems share its features, and all of them are able to cut, copy, and paste DNA. Historically, the discovery of such programmable systems has helped revolutionize medicine. CRISPR, for instance, is now the foundation of genetic medicines. But it will take much more work to learn what this system does, and whether it can be put to similar use.
Read more: anthropic.com/news/claude-di…
Unfortunately, a bunch of gatekeepers has succeeded in slowing down scientific progress.
We’re working with an independent advisory group of mathematicians to help OpenAI responsibly share advances in AI and mathematics.
The group will advise on how we assess and communicate new mathematical results, uphold academic and professional standards, and build tools that support mathematical research and learning.
Through this work, we want mathematicians to be at the center of shaping how AI supports mathematical understanding and how its benefits reach the wider community.
openai.com/index/advisory-gr…
For the investors among you - I’ve launched a free Telegram channel for each of leading 15 US stocks.
An alert the moment it crosses a whole percent. Plus the price at the open, midday and close.
No tips. No predictions. Just the price.
ylevels.com/channels/?v=2
95















