AlphaProof: olympiad-level formal reasoning (IMO 2024 silver)
Status
Peer reviewed
Recorded on · nature.com
- Announced
- Recordedpeer reviewed
What this status means
Source reading : Source read word for word (level 3 of 3)
Published in a journal or reviewed by named experts.
The gist
Google DeepMind published AlphaProof in Nature on 12 November 2025: a system that proves statements in Lean using reinforcement learning, reaching a silver-medal level at IMO 2024. Status: peer reviewed.
Cite this entry
A single status snapshot since Oct 11, 2026: a citation always points to a dated snapshot.
Cite snapshot 1, Oct 11, 2026
APA citation with the dated link
BibTeX @misc, key discoveria-alphaproof-nature-2025-11-12-r1
@misc{discoveria-alphaproof-nature-2025-11-12-r1,
author = {{DiscoverIA}},
title = {AlphaProof: olympiad-level formal reasoning (IMO 2024 silver)},
howpublished = {DiscoverIA registry, entry alphaproof-nature-2025-11-12},
note = {Status « Peer reviewed », snapshot 1, recorded on October 11, 2026. competition},
year = {2026},
url = {https://discoveria.ai/en/fiche/alphaproof-nature-2025-11-12/#releve-1},
urldate = {2026-10-11}
}Dated link opens this snapshot in the history
Text written by an AI (DiscoverIA pipeline), on Reviewed by: nobody yetReport an error
What has not been reviewed?
The status and its sources come from the registry; only the explanatory text has not been reviewed. A reported error is corrected, then the text is reviewed.
The status and its sources come from the registry; only the explanatory text has not been reviewed. A reported error is corrected, then the text is reviewed.
Key facts
- Claimed by
- Google DeepMind
announced on Nov 12, 2025 - Nature
- competition
- Field
- Mathematics (formal proof (Lean))
- Checked by
- official IMO 2024 marking (via the press, seen in a search-engine snippet)
- Sources (domains)
- 1 : 1 links only
- This entry
- AI draft, not reviewed
To understand
Analogy
It is like learning chess by playing against yourself again and again, except that each move is a proof step that an automatic checker accepts or rejects.
What it changes
Lean is a language in which a machine checks every proof. If the result holds, a system could produce olympiad-level proofs that can be checked mechanically, not just plausible answers.
What remains unclear
The silver level at IMO 2024 comes from an official evaluation reported by the press, seen only in an excerpt. No third-party verifier is named and no reproduction is listed.
1 source.
For experts
Sources(domains)
- primairenature.comlink only, reading not recorded
Method
Status “peer reviewed” recorded on 11 October 2026 from nature.com (page read, quotation checked word for word).
Limits
A Lean proof supplied by the claimant does not count as formal verification by a third party.
It is like learning chess by playing against yourself again and again, except that each move is a proof step that an automatic checker accepts or rejects.
Status history
Only one observation so far.
An observation is added; none is ever erased.
Report an error or suggest a source
Thank you, your message was sent.
Sending failed. Try again, or write to [email protected].