Tao, Gómez-Serrano, Georgiev and Wagner test AlphaEvolve on 67 maths problems
In 'Mathematical exploration and discovery at scale' (arXiv 2511.02864), Bogdan Georgiev, Javier Gómez-Serrano, Terence Tao and Adam Zsolt Wagner ran AlphaEvolve on 67 problems in analysis, combinatorics, geometry and number theory. It rediscovered the best known constructions in most cases and improved several. Some runs were chained with Deep Think and AlphaProof to produce proofs.
Key facts
- 67 problems; a public repository with per-problem notebooks
- Rediscovered state-of-the-art constructions in most cases and improved on several
- Pipeline: AlphaEvolve (constructions) → Deep Think (informal proof) → AlphaProof (formal proof) in some cases
- Tao blog post, 5 Nov 2025
Science result
- Field
- mathematics / analysis / combinatorics / geometry
- Problem
- Broad battery of optimisation-type open problems (e.g. inequalities, packings, finite-field Kakeya-type constructions)
- Result
- Systematic evidence that LLM-driven evolutionary search matches or beats best-known constructions across dozens of problems.
- AI system
- AlphaEvolve, Gemini Deep Think, AlphaProof
- Human role
- Human-led with AI tools: mathematicians chose problems and scorers
- Verification
- Constructions verifiable; preprint
- Status
- confirmed
What happened
Leading mathematicians stress-tested AlphaEvolve on a large, varied problem set and published both the successes and the failures.
Why it matters
Coming from Tao, it gave the maths community a credible, balanced picture of what AI search could do, just before the 2026 surge.
Changelog
- 2026-09-29: created
Related events
- AlphaEvolve: Gemini-powered agent discovers new algorithms ★★★★
- Adam Zsolt Wagner uses reinforcement learning to find counterexamples to open graph-theory conjectures ★★
Sources (3)
- paperMathematical exploration and discovery at scale (arXiv 2511.02864)
- discussionTerence Tao: Mathematical exploration and discovery at scale
- codeGitHub: alphaevolve_repository_of_problems
id: 2025-11-05-alphaevolve-tao-67-problems · updated 2026-09-29 · open in the interactive timeline