Demis Hassabis (@demishassabis)

AlphaEvolve가 자체적으로 탐색 절차(search procedures)를 발견해 5개의 고전적인 Ramsey 수에 대한 상계(bound)를 개선했으며, 일부는 10년 이상 만의 첫 개선이라는 성과를 냈습니다. 이는 수학 분야에서 AI 활용의 중요한 이정표로 평가됩니다.

https://x.com/demishassabis/status/2032267485735460867

#alphaevolve #ramsey #ai #mathematics

Demis Hassabis (@demishassabis) on X

Ramsey numbers are notoriously hard. Amazing to see AlphaEvolve improve bounds for 5 classical Ramsey numbers - some for the first time in 10+ years - by discovering search procedures itself. A big milestone in AI for maths - congrats to the team!

X (formerly Twitter)

Pushmeet Kohli (@pushmeet)

Google DeepMind의 AlphaEvolve가 수학 분야 AI 연구 진전을 공유했습니다. 극단 조합학(extremal combinatorics)에서 AlphaEvolve는 다섯 개의 고전적 Ramsey 수에 대한 새로운 하한을 확립하는 데 기여했으며, 이는 에르되시(Paul Erdős)도 어려움을 언급할 정도로 난이도가 높은 문제입니다. AI를 이용한 계산적 접근으로 학문적 성과를 도출한 발표입니다.

https://x.com/pushmeet/status/2031727892346941499

#googledeepmind #alphaevolve #ai #ramsey

Pushmeet Kohli (@pushmeet) on X

Happy to share new progress in AI for Maths @GoogleDeepMind . In extremal combinatorics, AlphaEvolve has helped establish new lower bounds for FIVE classical Ramsey numbers - a problem so challenging that even Erdős commented on its difficulty. Historically, computationally

X (formerly Twitter)

#Mathematics in the age of #ai

Will #deepmind #alphaevolve do for mathematics, what #alphafold did for proteomics?

Are LLMs driving Automated Theorem Provers the future of math?

https://youtube.com/watch?v=mS9Lr43cIB4

#fieldsmedal #lecture

Terence Tao - Mathematics in the Age of AI

YouTube

Google DeepMind veröffentlicht AlphaEvolve.

Das System nutzt LLMs und evolutionäre Ansätze, um Algorithmen für Multi-Agent Reinforcement Learning zu entwickeln. Dabei entstanden VAD-CFR und SHOR-PSRO.

Diese Methoden übertreffen menschliche Standards durch mathematische Strukturen, die für Entwickler kontraintuitiv wirken. Das System schreibt Logik neu. #AlphaEvolve #DeepMind #KI
https://www.all-ai.de/news/beitrage2026/alpha-evolve-deepmind

So übertrifft AlphaEvolve von DeepMind menschliche Programmierer

AlphaEvolve automatisiert das Design von Algorithmen durch evolutionäre Prozesse. Die entdeckten Methoden VAD-CFR und SHOR-PSRO setzen neue Maßstäbe.

All-AI.de
AI Transforms Scientific Discovery: How AlphaFold and AI Co-Scientist Are Reshaping Research

From solving the 50-year protein folding problem to generating research hypotheses in hours, AI tools like AlphaFold and Google's AI Co-Scientist are compressing discovery timelines from years to days

TechLife

#GoogleCloud announced the private preview of #AlphaEvolve!

A Gemini-powered coding agent, AlphaEvolve is designed to discover & optimize algorithms for complex engineering and scientific challenges - especially where brute-force or manual approaches break down due to massive search spaces.

👉 Learn more: https://bit.ly/48LxdJi

#InfoQ #AI #LLMs #Gemini

🧠 In diverse occasioni mi è capitato di parlare del progetto #AlphaEvolve di #Google… e ora sta arrivando su Google Cloud.
👉 I dettagli: https://www.linkedin.com/posts/alessiopomaro_alphaevolve-google-gemini-activity-7404492798437068801-Ncgv

___  
✉️ 𝗦𝗲 𝘃𝘂𝗼𝗶 𝗿𝗶𝗺𝗮𝗻𝗲𝗿𝗲 𝗮𝗴𝗴𝗶𝗼𝗿𝗻𝗮𝘁𝗼/𝗮 𝘀𝘂 𝗾𝘂𝗲𝘀𝘁𝗲 𝘁𝗲𝗺𝗮𝘁𝗶𝗰𝗵𝗲, 𝗶𝘀𝗰𝗿𝗶𝘃𝗶𝘁𝗶 𝗮𝗹𝗹𝗮 𝗺𝗶𝗮 𝗻𝗲𝘄𝘀𝗹𝗲𝘁𝘁𝗲𝗿: https://bit.ly/newsletter-alessiopomaro 

#AI #GenAI #GenerativeAI #IntelligenzaArtificiale #LLM 

The story of Erdős problem #126

Problem 1026 on the Erdős problem web site recently got solved through an interesting combination of existing literature, online collaboration, and AI tools. The purpose of this blog post is to try…

What's new

필즈상 수상자가 AI로 수학 문제 67개 풀어본 결과: 300년 난제도 개선

필즈상 수상자 테렌스 타오가 AI 도구 AlphaEvolve로 67개 수학 문제를 테스트한 결과. 75% 재발견, 20% 개선, 300년 묵은 키싱 넘버 문제도 돌파했습니다.

https://aisparkup.com/posts/6365

Readings shared November 7, 2025

The readings shared in Bluesky on 7 November 2025 are: An introduction to formal real analysis (Lecture 16: Completeness of the real numbers). ~ Alex Kontorovich. #ITP #LeanProver #Math The path to a

Vestigium