💡 Navier-Stokes lost in translation: Why Lean verification of AI autoformalisation does not guarantee correct natural language proofs arxiv.org/abs/2610.08144
@jbzfn.bsky.social
@jbzfn.bsky.social💡 Navier-Stokes lost in translation: Why Lean verification of AI autoformalisation does not guarantee correct natural language proofs arxiv.org/abs/2610.08144
@donwebmedia.bsky.socialTerence Tao responde al volcado matemático de OpenAI
Terence Tao OpenAI matemática: repasá qué criticó, qué es Math 1.0 y qué demostraciones siguen sin verificar tras el volcado del 6 de octubre
#terencetao #openai #matemática #math10 #lean
@mathify-dev.bsky.socialOpenAI’s New Matrix Multiplication Bound: ω ≤ 2.25
One result family in OpenAI’s October 2026 mathematics release concerns the exponent ω of matrix multiplication.
@pulseofnations.lolOpenAI released 377 new mathematical results spanning algebra, number theory and topology, verified with the Lean proof assistant, but mathematicians question the process and the model behind them.
#AdvisoryBoard #Lean #Mathematics #NavierStokes #OpenAI #Research
@mathify-dev.bsky.socialOpenAI’s Unique Games Claim Is Not P vs NP
A manuscript in OpenAI’s October 2026 mathematics release tackles Unique Games, a conjecture open for more than two decades.
@waynerad.bsky.socialOpenAI has released 722 math papers "covering 372 result families" of previously unsolved problems in mathematics. Crucially not all of these have computer-verifiable proofs in Lean.
#solidstatelife #ai #genai #codingai #mathematics #proofs #lean
@fabmusacchio.bsky.socialThe authors find exactly this in #OpenAI’s recent #NavierStokes work: Different bounds, different proof strategies, but still valid #Lean code.
2/N
@fabmusacchio.bsky.social#NaviesStokes lost in translations: #Lean can verify a formal proof perfectly well, while the #AI may have changed the actual #mathematical argument during translation
1/N
@3939ai.bsky.socialOpenAI mathの難問「ガウスの堀」を題材に、Lean 4形式検証から3Dサイエンスアートへ直結するパイプラインを構築！
基礎補題をLeanでコンパイル（.olean生成）してGate通過→有限素数グラフのBFS探索→WebGLで可視化🌌
#Lean #FormalVerification #OpenAI #数学 #可視化 #HTML #JavaScript
@fmontesi.bsky.socialThe framework now supports, among other things, basic modal logic, the modal cube, basic temporal logic, and Hennessy–Milner Logic, together with reusable metatheory and automation.
Preprint: www.fabriziomontesi.com/publication/...
#CSLib #Lean #FormalMethods #FORM
5/5
@mathify-dev.bsky.socialWhy 7/8 Matters in OpenAI’s Zeta-Function Claim
Follow for the equations hiding behind the news.
@mathify-dev.bsky.socialOpenAI’s Navier–Stokes Singularity Claim, Explained
Follow for the equations hiding behind the news.
@pulseofnations.lolOpenAI released 722 manuscripts from an unreleased frontier model covering hundreds of open math problems, with proofs formalized in Lean. Verification and credit questions now land on human referees.
@gasmholic.bsky.social✨ 🍽️ [맛집 사장님 단골마트 가즘홀릭] 식자재마트 가즘홀릭 단독 제안
"이른 새벽에 발로뛰며 대학생과 함께 하는 #Lean #Start-up 다용도 향신료 그리고 요링에 빠지지 않는 #팔각 #팔각향 [500g]"
💬 후기: 포장 꼼꼼하고 배송도 빠르네요. 맛은 말할 것도 없이 일품입니다.
🔥 특가: 76% (32,900원 -> 7,890원)
👉 구경가기: smartstore.naver.com/gasmholic/pr...
@mathify-dev.bsky.social722 Manuscripts, 372 Families—and a Verification Bottleneck
Follow for the equations hiding behind the news.
#mathematics #formalverification #lean #artificialintelligence
@mathify-dev.bsky.social722 Manuscripts, 372 Families—and a Verification Bottleneck
Follow for the equations hiding behind the news.
#mathematics #formalverification #lean #artificialintelligence
@gradientbrief.bsky.socialAIProver is an agentic framework that combines a 119B open-weight language model with an evolved tool-calling harness for proof auto-formalization and synthesis, using verifiers to address missing concepts and…
#AIResearch #AutoFormalization #Lean #MathAI
https://arxiv.org/abs/2610.05367
@harushark3.bsky.social