Claim from OpenAI
The unit group L_F2(1,2)× of the binary Leavitt algebra is not sofic.
Quoted from the source · OpenAI · 2026-08-01, p.78
Author involvement
Compiled from OpenAI's public release, manuscript, and repository. OpenAI did not participate in or confirm this Significance record.
Summary · for readers
- In one sentence
- OpenAI says one specific infinite group — the unit group of a small algebra built from two symbols — cannot be approximated by finite shuffles. That property is called soficity, and the manuscript states this group does not have it.
- Claim
- OpenAI affirme qu'un groupe infini précis — le groupe des éléments inversibles d'une petite algèbre construite à partir de deux symboles — ne peut pas être approché par des permutations finies. Cette propriété s'appelle la soficité, et le manuscrit soutient que ce groupe ne la possède pas.
- Checked
- Significance a reconstruit le code Lean publié à un commit fixé et a enregistré que la compilation et la vérification des axiomes sont allées à leur terme, les deux journaux étant empreintés. Le fichier du manuscrit a été téléchargé et empreinté. Ce sont des vérifications portant sur des artefacts, pas sur les mathématiques.
- Not checked
- Personne n'a établi le lien entre le théorème 1.1 du manuscrit et l'énoncé que contient réellement le code Lean ; la cible formelle du dépôt est un énoncé d'existence plus large. Aucune évaluation par un mathématicien n'est consignée ici. Les trois invitations ouvertes désignent ce travail.
- Claim
- تؤكد OpenAI أن زمرة لا نهائية محددة — زمرة العناصر القابلة للعكس في جبر صغير مبني من رمزين — لا يمكن تقريبها بتباديل منتهية. تسمّى هذه الخاصية الصوفية، ويذهب المخطوط إلى أن هذه الزمرة لا تتمتع بها.
- Checked
- أعادت Significance بناء شيفرة Lean المنشورة عند إصدار مثبّت، وسجّلت أن عملية البناء وفحص البديهيات اكتملتا، مع بصمة رقمية لكلا السجلّين. كما جرى تنزيل ملف المخطوط وحساب بصمته. هذه فحوص على المصنوعات الرقمية لا على الرياضيات نفسها.
- Not checked
- لم يتتبع أحد الصلة بين المبرهنة 1.1 في المخطوط والعبارة التي تتضمنها شيفرة Lean فعلياً؛ فالهدف الصوري في المستودع عبارة وجود أوسع. ولا يتضمن هذا السجل أي تقييم من رياضي. والدعوات المفتوحة الثلاث تشير إلى هذا العمل.
Start here · for reviewers
Main deduction
Start with the manuscript's Theorem 1.1 and the companion Comparator target, then trace whether they state the same mathematical result.
Risk points
The central risk is the gap between the manuscript's concrete unit-group theorem and the formal target's broader finitely-presented existence statement.
The proof's use of property (T), Kun's theorem, and the expander decomposition is the main mathematical reading target after correspondence.
Useful background
Basic soficity, property (T), expander graphs, and the role of a finitely presented group presentation.
Needs checking
Compare the theorem represented by the Lean/Comparator target with the concrete group named in the manuscript.
- Suggest another focused check
Formalization handoff
A compact map for someone who wants to formalize this claim. It records preparation and open work, not a mathematical verdict.
- Target
- The manuscript's Theorem 1.1 and the Comparator target for a finitely presented non-sofic group. Significance’s interpretation · Significance editor · 2026-08-20
- System
- Lean 4 with the repository's Comparator challenge Significance’s interpretation · Significance editor · 2026-08-20
- Work state
- Artifact reproduced
- Code
- formalization repository · commit
c510a55434c0…
Definitions
Soficity, the concrete unit group named in the manuscript, and finitely presented group must be compared before treating the paper and code as the same target.
Prerequisites
Basic group presentations, sofic approximations, property (T), and the repository's Comparator workflow.
Open questions
Trace the declaration chain and record exactly where the concrete Leavitt-algebra statement does or does not meet the broader formal target.
If extending the formalization, identify the smallest missing declaration rather than attempting the whole manuscript at once.
Review activity
- Independent reruns
- 1
- Math assessments
- 0
- Written reviews
- 0
- Open checks
- 2
What Significance checked: Significance rebuilt the published Lean code at a pinned commit and recorded that the build and the axiom check ran to completion, with both logs hashed. The manuscript file was downloaded and hashed. These are checks on artifacts, not on the mathematics.
Still open: Nobody has traced the manuscript's Theorem 1.1 to the statement the Lean code actually contains; the repository's formal target is a broader existence statement. No mathematician's assessment of the argument is recorded here. The three open invitations name this work.
Careful wording
OpenAI has published a claimed proof with a Lean formalisation checked by its own pipeline; the link between the paper's theorem and the formal statement has not been traced by anyone else, and no independent mathematical assessment is on record.
Copy for sharing
Copy and paste this summary.
OpenAI has published a claimed proof with a Lean formalisation checked by its own pipeline; the link between the paper's theorem and the formal statement has not been traced by anyone else, and no independent mathematical assessment is on record. Checked: Significance rebuilt the published Lean code at a pinned commit and recorded that the build and the axiom check ran to completion, with both logs hashed. The manuscript file was downloaded and hashed. These are checks on artifacts, not on the mathematics. Not checked: Nobody has traced the manuscript's Theorem 1.1 to the statement the Lean code actually contains; the repository's formal target is a broader existence statement. No mathematician's assessment of the argument is recorded here. The three open invitations name this work. As of 2026-08-01T14:09:51Z (freshness: current) Full record: https://hjyuh.github.io/significance/2026-openai-nonsofic-groups/ Significance records evidence. It does not judge the mathematics.
- Scope
- The manuscript identifies a concrete countable unit group; its public Comparator target separately states existence of a finitely presented non-sofic group. Significance’s interpretation · Significance editor · 2026-08-01
- Excluded
-
The manuscript says this result does not determine whether the unit group is hyperlinear. Quoted from the source · OpenAI · 2026-08-01, p.80
- Source
- OpenAI, Ten Advances in Mathematics and Theoretical Computer Science
Evidence — 2 entries
-
Formal code published by the source
ev-openai-lean-artifactPublished by the source. Any independent rerun appears as a separate entry.
OpenAI published NonSoficGroup.lean and a Comparator challenge targeting existence of a finitely presented non-sofic group. At the time of this source-publication entry, Significance had not yet attached an independent execution receipt; the later formal-artifact entry records that reproduction.
-
Independent formal-code run
ev-lean-c510a55434c0… · execution record attachedStatement in the paper
The unit group L_F2(1,2)× of the binary Leavitt algebra is not sofic.
Statement represented in lean4
Open link @
c510a55434c0…At the pinned commit, ComparatorChallenges/D_NonSoficGroup.json maps solution module NonSoficGroup to SoficGroups.SourceTopLevelCompressionFinal.exists_finitelyPresented_nonsofic_group. This identifies the declaration the repository asks Comparator to check; it does not independently establish correspondence with the manuscript's concrete Leavitt-algebra theorem.
Allowed assumptionsStandard classical Lean profile —propext,Classical.choice,Quot.soundCode buildCompleted · 2026-08-02 · log fingerprintd0eb06613908…Assumption checkCompleted · 2026-08-02 · log fingerprintd0eb06613908…Technical execution details
Environmentsignificance-lean 0.1.0 · image fingerprintsha256:e18db1519e04…· Significance automated check
Plain-language explanation
A group here is a set of reversible operations you can undo and combine. Soficity asks whether an infinite group can be imitated, to any accuracy you like, by shuffling a finite deck of cards: every product you can form in the group is matched by a shuffle that agrees almost everywhere. The manuscript names the reversible elements of a small algebra built from two symbols. Its claim is that no finite shuffle imitates this group well enough. What Significance holds is the artifacts around that claim -- the manuscript file, the formal code, the logs of rebuilding it -- not a reading of the argument itself.
Mathematical context
The manuscript proposes an explicit boundary for finite-permutation approximation: a concrete group built from the binary Leavitt algebra that is claimed not to admit sofic approximations. The public Lean target records a related existential finite-presentation statement, so correspondence between the concrete and formal claims remains a useful, separate audit task.
Importance
The announcement describes the result as addressing a central open question in group theory. Quoted from the source · OpenAI
Limits
- This record does not issue a whole-paper correctness judgement.
- Correspondence between formal and informal statements is attested, not machine-checked; a formalization may omit informal hypotheses.
- The formal adapter performs narrow artifact checks. It is not a general mathematical verifier.
- Open invitations identify work not yet represented by a completed evidence entry.