Jorge Hernandez 🇺🇦 🏳️‍🌈 (@braneloop) profile picture

Principal ML Engineer • AuDHD • Tweets: ML/AI, Math, Neuroscience, Physics, Philosophy, other stuff.

In transit ...
Joined October 2015
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
Replying to @yifanzhang_
Translation, for the hoopers in the comments:
16
21
5
693
162,834
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
JUST IN: Riemann Hypothesis has been solved. alpo.ge/remark.pdf
1,301
4,738
3,500
37,104
5,634,285
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
From Ravi Vakil at the @amermathsoc ECBT meeting: "Alessandro Masullo has a proposed proof of GRH, and Brian Conrad has been looking closely at it for months and gave extensive feedback which is being incorporated, and it is so far looking good. It is all human, with no AI."
9
44
16
307
25,867
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
This is massive. A story I've been following for years. He thought he had more time but didn't. Alessandro Masullo announces a separate, purely human proof of the Riemann hypothesis. math216.wordpress.com/2026/1…
51
222
76
2,342
122,721
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
Summary of the RH proof, courtesy of Chris Birkbeck 1. RH is deduced from "no Dirichlet L-function vanishes for Re(s) > 1/2". This step is short and standard. 2. That nonvanishing is deduced from three things: a growth bound on certain Galois cohomology groups (they look like Tate–Shafarevich groups) of a family of subfields of cyclotomic fields; a covolume bound coming from arithmetic geometry; and statements about theta integrals on unitary groups, together with local Langlands for GL_n. I have not yet understood how these combine to exclude zeros. 3. The cohomology bound is proved by what looks like an automorphy-lifting argument. It genuinely uses the existence of Galois representations attached to regular algebraic, polarized cuspidal automorphic representations of GL_n over CM fields, along with base change, Jacquet–Shalika, Clozel's purity, Bernstein–Zelevinsky, the main theorem of CM, Deligne–Ribet and more. 4. Those Galois representations are not assumed. The project derives them from a geometric construction for unitary groups (Eichler–Shimura relation, crystalline above l) plus theta lifts between unitary groups. As far as I can tell this is not the route in the literature. 5. The arithmetic geometry side lists, by name, resolution of singularities in characteristic zero, arithmetic Hilbert–Samuel, Moriwaki's bigness theorem, toroidal compactifications of Siegel varieties, Kawamata–Viehweg and Nadel vanishing, Grothendieck–Messing and prismatic Dieudonné theory. There are no axioms, so each of these has to be proved inside the project. leanprover.zulipchat.com/#na…
1
24
3
149
12,919
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
Lean experts are currently investigating the validity of the claimed formalised proof of the Riemann Hypothesis - follow along at leanprover.zulipchat.com/#na…
16
135
18
1,079
95,554
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
Roughly from a month ago. terrytao.wordpress.com/2026/…
8
45
6
616
37,898
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
By combining Antithesis with formal methods, we can mathematically prove safety for conformant models of production distributed systems! We demonstrated this process in depth for etcd-raft and HashiCorp Raft. For a substantial subset of their features, we can show how, after patching a few bugs, their implementations achieve the properties stated in the Raft paper, such as Election Safety and State-Machine Safety.
6
14
3
84
7,284
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
Levent Alpoge @__alpoge__ claims to have proved RH and GRH with Claude (!!) alpo.ge/remark.pdf
32
134
65
1,210
141,307
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
1) The formalized statement in Levent's drop is in fact GRH. 2) Still compiling, will take a few hours, but c'mon, it's gonna compile, obv that was checked ahead of time 3) Technically it could be a Lean bug but frontier models virtually never do that on accident. <<1% it's a bug
37
89
20
2,232
132,936
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
1/5 EBMs are usually taught as one model family among many. In our new tutorial with @du_yilun, we use them as a single lens on machine learning. "An Energy-Based Perspective on Machine Learning" arxiv.org/abs/2611.14383
3
34
7
220
22,011
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
犬を抱っこするアザラシと、 犬に抱っこされるアザラシを描きました
15
314
12
3,581
28,864
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
I find it impossible to use Gemini for maths... like the page just constantly freezes for a solid few minutes
34
7
2
312
17,963
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
It was pointed out to me offline that so many of the recent big recent breakthroughs were “Rutgers” theorems: - Quasi-RH is basically: take Iwaniec’s life work (catching Mobius with a root number of Gauss sums, amplification, deforming in families, Poisson, Voronoi, large sieve,…) and have AI smash it all together. - Classification of finite simple groups (Rutgers: Gorenstein et al) - Yang-Mills: Tadeusz Balaban (Rutgers; rigorous renormalization-group, lattice gauge theory…) Obviously I’m exaggerating, and many others contributed everywhere. But way to go Rutgers!
12
15
1
208
18,194
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
I’ll see you at CoevCon in two weeks. Btw moots you are forewarned that when I say I’m six feet tall I mean I’m six feet tall like actually. coevcon.com/
3
1
21
482
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
Milky Way vs Northern Lights - what a light show! Spectacular image by Guillaume Scheib, taken about a month ago from Kvaløya in northern Norway. instagram.com/guillaume_sche…
78
573
35
3,854
63,062
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
I hear SGA VI just dropped and so many people are going thru it. Grothendieck would be proud
2
3
1
91
5,719
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
Each word you say is planned by a small patch of brain above your left ear, about a quarter of a second before you say it. It sits in a temporal lobe region textbooks describe as only for listening. Our paper today in @ScienceMagazine: science.org/doi/10.1126/scie… @UCSF @NeurosurgUCSF
174
1,370
151
8,834
959,403
Jorge Hernandez 🇺🇦 🏳️‍🌈 retweeted
Frobenius structures! (aka the most beautiful things in math and physics)
4
49
2
478
19,474