Anier Velasco Sotomayor

Asked myself ‘are all proofs by contradiction the same?' and few mins later I'm staring at universe hierarchies. Every paradox is the same diagonal trick: type theory exists bc someone said 'make that a type error.' Feels nice re-deriving DTT from a simple question.

Arithmetic. Vision. Navigation. Planning. What if they're all the same task? We introduce group composition as a unifying abstraction for many learning problems & show neural networks crack it using Fourier! Led by @KuninDaniel @giovannimarchet @AdeleMyersPhD @hopfbifurcator🌟

@Anthony_Bonato I'd reply: Before we answer...what standard of certification do you accept: symbolic proof, formally verified computation, just search?

Book giveaway! I'm giving away a copy of The Sky Took Us. Must have DMs open + mailing address. “If aliens come to Earth and demand the value of the Ramsey number R(5), what would you do?” Best reply wins. Keep it clean :)

oh, did i say chapter? i meant _chapters_ We've just released draft Chapter 6 (Grids) and Chapter 7 (Group Convolutions on Homogeneous Spaces) of the GDL Book Alice's journeys in geometric wonderland continue #️⃣🌍

@PetarV_93 @TacoCohen I was about to email you guys! haha

Formalizes mathematics textbooks in Lean via multi-agent collaboration github.com/facebookresear…

Today, at the @DARPA expMath kickoff, we launched 𝗢𝗽𝗲𝗻𝗚𝗮𝘂𝘀𝘀, an open source and state of the art autoformalization agent harness for developers and practitioners to accelerate progress at the frontier. It is stronger, faster, and more cost-efficient than off-the-shelf alternatives. On FormalQualBench, running with a 4-hour timeout, it beats @HarmonicMath's Aristotle agent with no time limit. Users of OpenGauss can interact with it as much or as little as they want, can easily manage many subagents working in parallel, and can extend / modify / introspect OpenGauss because it is permissively open-source. OpenGauss was developed in close collaboration with maintainers of leading open-source AI tooling for Lean. Read the report and try it out:

One family. Generations of explorers. The Baldwins’ legacy reaches far beyond Earth.

@miniapeur Enumerative combinatorics

What is your favourite topic in mathematics?

repo: github.com/aniervs/microg…

Built a micrograd-style autodiff engine in Lean 4 (@leanprover): tape-based backprop, trains on MNIST, and the derivative rules are formally proven correct by Lean's kernel. Inspired by @karpathy's micrograd.

@Golovanov_ammoc I think LaTeX was made by Leslie Lamport, on top of TeX (by Knuth).

This is what I wrote in the first week, when I learnt LaTeX. I can’t tell how overwhelmingly happy I felt to see myself in position to write classical typewriter era document on my own. I felt very grateful towards Donald Knuth, who created this wonderful platform.

@_init_politics tampoco así...por ejemplo INF y -INF son elementos de los reales extendidos (en.wikipedia.org/wiki/Extended_…), a aparecen en en semianillos tropicales. Entonces, si son números. Lo que cambia es la definición de "número".

El infinito no es un número , es un concepto sino inf/inf fuera igual a 1, pero no, eso es cualquier cosa.

UlamAI Prover - an open source Theorem Prover and Formalizer in Lean 4 We've just released a CLI tool that connects with your Codex/Claude Code/Ollama and allows you to prove and formalize mathematics in Lean directly from .tex files. The goal is to make progress in mathematics faster and more reliable.

@Golovanov_ammoc Planning to write anything on how you work internally, with the hope of inspiring people in other countries to run similar programs?

This sheds some light about admissions at AMMOC.

Statistical Learning Theory in Lean 4: Empirical Processes from Scratch arxiv.org/abs/2602.02285

Excited to share our latest work: "Semi-Autonomous Mathematics Discovery with Gemini." We used Gemini to systematically evaluate 700 "open" conjectures in the Erdős Problems database. The result? We addressed 13 problems marked as open—finding 5 novel autonomous solutions and identifying 8 existing solutions missed by previous literature. Read the full case study here: arxiv.org/abs/2601.22401

@Arciel___Baez @dianalokada talking about two different statements: OP: for every x (x good => x starts with c) yours is a counterexample for the converse implication :)

@dianalokada Communism?😂

all good things in life start with C: Cash Claude Creatine Celsius Codex Coffee

@Arciel___Baez @d4rsh_tw What’s wrong with having a mind like Einstein’s?

@d4rsh_tw I don't think having the mind of Einstein would be enjoyable. Also when you have the money of Elon I don't think you want that popularity. I think I would take just the money. If I had to pick a second one I would go for the looks but I'm pretty fine with mine.

You can only choose any two

BREAKING: Logical Intelligence has signed Meta’s former Chief AI Scientist @ylecun as the Founding Chair of the Technical Research Board