19 passages checked, 34 Lean locations (and 6 related declarations).
Central statements: 9 of the 19 passages state the article's subject or a named result or definition; the other 10 are supporting (examples, setup). Of these 9: ✓6 formalized,◔3 partial.
✓ Formalized a Lean declaration states it, proved, or a Lean definition defines it
◐ Statement only Lean states it, but the proof is not finished
◔ Partial Lean states part of it, or a more general or special form
✕ Not formalized searched every library and not found
Libraries searched
Searched 48 Lean libraries, pinned 2026-10-17, 3 left out (library set u1).
How these links were made
Accuracy, from an audit of a sample of 224 items (blind, not independent): Strong links (exact, close or broad match) on supporting passages that have exactly the right relation: 92% (95% interval 49–99%), from 36 drawn of 1029, weighted by relation and library, an effective sample of 5.1; Links in a counting relation (exact, close, broad, narrow match or part) on supporting passages that have exactly the right relation: 73% (95% interval 40–92%), from 48 drawn of 1740, weighted by relation and library, an effective sample of 8.1; Central and subject passages whose formalization status is right: 86% (95% interval 73–94%), from 44 drawn of 1905, weighted by recorded status, an effective sample of 42; 'Not formalized' on supporting passages that is right (no declaration in the 48 libraries states it): 60% (95% interval 25–87%), from 12 drawn of 1236, weighted by article, an effective sample of 6.1.
Audit run tierA-batch1-audit, run by jack, 2 Nov 2026, 21:10 – 3 Nov 2026, 02:10 UTC: judged by Claude Opus 5.5 (claude-opus-5-5).
Record set tierA-b1b (cf-locate-recordset/2, statuses by status rule 3), exported 1 Nov 2026 by formal-companion at commit d85385a.
Seal sha256 23710337d053e8212378a9e580805628bcdfda61a6fc431e65f24a4a16c9e602.
Inventory run inv-tierA-b1r, run by jack, 30 Oct 2026, 07:23–20:00 UTC: worker Claude Opus 5 (claude-opus-5), reviewer Claude Opus 5 (claude-opus-5). It listed the formal passages, each with its kind and prominence, that the locate run then searched.
Locate run loc-tierA-b1b, run by jack, 31 Oct 2026, 04:59 – 1 Nov 2026, 03:06 UTC: worker Claude Opus 5 (claude-opus-5), reviewer Claude Opus 5 (claude-opus-5).
Challenge run chal-tierA-b1, run by jack, 1 Nov 2026, 03:19–14:25 UTC: worker Claude Opus 5 (claude-opus-5), reviewer Claude Opus 5 (claude-opus-5). In this article it re-checked every Lean location on a central or subject passage (22 locations); on supporting passages, a seeded sample (seed 0; drawn at 0.5 for strong matches, 0.2 for weaker links and 0.25 for absences): 9 of 12 locations and 0 of 3 absences. In all, 31 of 34 Lean locations (91%) and 0 of 3 absences (0%) were re-checked.
Passage text: 19 of 19 passages shown in full, each the slice of the pinned revision's text that its record's span sha256 hashes, checked when the store was built.
Lean code: each declaration's type, from the declaration corpus of universe u1 (built 18 Oct 2026, sha256 a364fd6aede5), at the commit its card links.
How the text is shown
Each passage is shown in full, as revision 1355154855 states it (checked against its record's span hash when the store was built). Where a passage is only part of a sentence, it is highlighted and the rest of the sentence is shown in grey around it, from the same copy of the revision; a list item also shows the sentence that introduces it. Formulas are drawn from their TeX; hover one to see it.
Each card's Lean code is its declaration's type as Lean prints it, from the declaration corpus of library set u1 at the card's commit, laid out as a header: binders before the colon, then the hypotheses as arrows (the corpus keeps no hypothesis names) and the conclusion, in Mathlib's notation (𝓝, π, R[X], Type*). It is not the source text; Source opens that.
Passages
What “exact match”, “more general” and the other labels mean
exact match states the passage exactly
close match states it up to a convention or a difference the note names
more general states a more general result, which implies it
special case states a special case of it
states part states one part of it
related states something related but different; never counted
Theorem · main statement
✓Formalized
Formalized: Mathlib states a more general result (proved).
Its correctness is a theorem of ZF, and relies on the fact that the ordinal numbers are well-ordered, and thus a statement that is not universally true for all ordinals must have a minimal counterexample.
Why more general: The library obtains the principle from the well-foundedness instance of the ordinals, which is the ground the passage names; the proof lives in Lean's type theory, so the particular axiom system the passage points at is not itself an object of the index.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, medium confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
How it differs: It registers the ordinals as a type with well-founded strict order, the fact the passage says the principle rests on.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 2 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Formalized: Mathlib states a more general result (proved).
Its correctness is a theorem of ZF, and relies on the fact that the ordinal numbers are well-ordered, and thus a statement that is not universally true for all ordinals must have a minimal counterexample.
Why more general: Mathlib states it for any preorder whose strict order is well-founded and expresses leastness with its minimality predicate; on the linearly ordered ordinals a minimal failing ordinal is a least one.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Why more general: The library phrases the conclusion through a nonempty set of elements rather than through a failing property.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 2 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Formalized: Mathlib states a more general result (proved).
In fact, this principle is also true for arbitrary well-ordered sets, but since any well-ordered set can be indexed by ordinals in an order-preserving way, it suffices to establish the principle for ordinals.
Why more general: Mathlib states it for any type whose strict order is well-founded, so the linearity carried by the passage's set is not required.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Why more general: The core library states it for an arbitrary relation rather than for an order.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 2 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit c9afcdd.
Why a special case: It is the passage's principle for the naturals alone.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 3 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Formalized: Mathlib states this exactly (a definition).
In fact, this principle is also true for arbitrary well-ordered sets, but since any well-ordered set can be indexed by ordinals in an order-preserving way, it suffices to establish the principle for ordinals.
Note: The ordinals below the order type form the initial segment the passage speaks of, and the map carries each of them to the element it indexes, in both directions respecting the order.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Why close, not exact: The library packages the comparison as an order embedding of the set into the ordinals whose image is cut off at a top ordinal, rather than as a map onto that segment.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 2 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Formalized: Mathlib states a more general result (proved).
Let P(α) be a property defined for all ordinals α. Suppose that whenever P(β) is true for all β < α , then P(α) is also true.[2] Then P is true for all ordinals.
Why more general: Mathlib states it for every type with well-founded strict order; the index holds no declaration of the principle written for the ordinals alone.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Why more general: The core library states induction along an arbitrary relation rather than along the order of the ordinals.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 2 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit c9afcdd.
Why more general: Mathlib writes minimality as the absence of a smaller failing ordinal, and takes its hypothesis as the existence of a failing ordinal where the passage says the property is not true of all ordinals; passing between the two readings is negation over the linear order of the ordinals.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, medium confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Why more general: The library states minimality as no member of the set lying below the chosen one, where the passage states that the property holds below it.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 2 for this passage, medium confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Why more general: The declaration rests on the well-foundedness instance for the ordinals, which is the ground the passage gives for the existence of such an ordinal.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
How it differs: It registers the well-founded strict order on the ordinals, which the passage names as the guarantee behind the minimal counterexample.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 2 for this passage, high confidence.
Not re-checked: sampled out of challenge run chal-tierA-b1 by jack, in which Claude Opus 5 (claude-opus-5) re-checked a sample on 1 Nov 2026 (UTC).
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Theorem · main statement · Overview › Induction by cases
◔Partial
Partial: Mathlib states a special case (a definition).
A proof by transfinite induction is often broken down into three cases: Zero case: Prove that P(0) is true. Successor case: Prove that for any successor ordinal α + 1, P(α + 1) follows from P(α) (and, if necessary, P(β) for all β < α). Limit case: Prove that for any limit ordinal λ, if P(β) holds for all β < λ, then P(λ).
Why a special case: Its successor hypothesis asks for the step from the value at the predecessor alone, so it covers those readings of the passage in which the values below the predecessor are not used.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Why a special case: It singles out the minimal element where the passage names zero, and its successor step likewise may use only the value at the predecessor.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 2 for this passage, medium confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
How it differs: It is the one-hypothesis form of the principle; getting the passage's three cases out of it needs the separate division of ordinals into zero, successors and limits.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 3 for this passage, medium confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Definition · main statement · Transfinite recursion
✓Formalized
Formalized: Mathlib states a more general result (a definition).
Transfinite recursion is similar to transfinite induction; however, instead of proving that something holds for all ordinal numbers, we construct a sequence of objects, one for each ordinal.
Why more general: Lean allows the object attached to an ordinal to live in a type depending on that ordinal, and the constant choice is the one the passage describes.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Why more general: It is stated for any type whose strict order is well founded, the ordinals being one such type.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 2 for this passage, medium confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Formalized: Mathlib states a more general result (a definition).
As an example, a basis for a (possibly infinite-dimensional) vector space can be created by starting with the empty set and for each ordinal α > 0 choosing a vector that is not in the span of the vectors {vβ ∣ β < α}.
def Module.Basis.ofVectorSpace(K : Type*)(V : Type*)[DivisionRing K][AddCommGroup V][_root_.Module K V] :Module.Basis (↑(Module.Basis.ofVectorSpaceIndex K V)) K V
Why more general: It is stated over a division ring rather than over a field, and the library obtains its index set from a maximal linearly independent set instead of choosing a vector at each stage.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Why more general: Stated over a division ring, it asserts that some set of vectors of the space carries a basis structure.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 2 for this passage, high confidence.
Not re-checked: sampled out of challenge run chal-tierA-b1 by jack, in which Claude Opus 5 (claude-opus-5) re-checked a sample on 1 Nov 2026 (UTC).
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Transfinite Recursion Theorem (version 1). Given a class function[3] G: V → V (where V is the class of all sets), there exists a unique transfinite sequence F: Ord → V (where Ord is the class of all ordinals) such that F(α) = G(F ↾ α) for all ordinals α, where ↾ denotes the restriction of F's domain to ordinals < α.
theorem WellFoundedLT.fix_eq{α : Type u}[LT α][WellFoundedLT α]{motive : α → Sort*}(ind : (x : α) → ((y : α) → y < x → motive y) → motive x)(x : α) :WellFoundedLT.fix ind x = ind x fun y x ↦ WellFoundedLT.fix ind y
Why only part: Lean hands the earlier values to the rule as a family indexed by the smaller ordinals, where the passage speaks of the restriction of F below the ordinal.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
theorem WellFounded.fix_eq{α : Sort u}{C : α → Sort v}{r : α → α → Prop}(hwf : WellFounded r)(F : (x : α) → ((y : α) → r y x → C y) → C x)(x : α) :hwf.fix F x = F x fun y x ↦ hwf.fix F y
Why only part: It is stated for an arbitrary well-founded relation on a type, with the well-foundedness proof an explicit argument.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 2 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit c9afcdd.
Why only part: It supplies the function the theorem asserts to exist, while the three conditions on that function are separate statements.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Why only part: It establishes the first of the three conditions for the constructed function, not the whole existence claim.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 2 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Why only part: It establishes the second of the three conditions for the constructed function only.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 3 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
theorem Ordinal.limitRecOn_limit{motive : Ordinal → Sort*}(o : Ordinal)(H₁ : motive 0)(H₂ : (o : Ordinal) → motive o → motive (o + 1))(H₃ : (o : Ordinal) → Order.IsSuccLimit o → ((o' : Ordinal) → o' < o → motive o') → motive o)(h : Order.IsSuccLimit o) :Ordinal.limitRecOn o H₁ H₂ H₃ = H₃ o h fun x _h ↦ Ordinal.limitRecOn x H₁ H₂ H₃
Why only part: It establishes the third of the three conditions for the constructed function only.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 4 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
at 0 the function takes the given set as its value
Ordinal.limitRecOn_succMathlib(proved)
at a successor ordinal the value is the second class function applied to the value at the predecessor, for every ordinal
Ordinal.limitRecOn_limitMathlib(proved)
at a nonzero limit ordinal the value is the third class function applied to the restriction of the function below that ordinal
Given a set g1, and class functions G2, G3, there exists a unique function F: Ord → V such that F(0) = g1, F(α + 1) = G2(F(α)), for all α ∈ Ord, F(λ) = G3(F ↾ λ), for all limit λ ≠ 0.
Why only part: Lean states it for an arbitrary family of target types and arbitrary case data, among which the passage's g1, G2 and G3 are one choice.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Why only part: Lean's successor datum may also depend on the ordinal itself, and the passage's G2 is the choice that ignores it.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 2 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
theorem Ordinal.limitRecOn_limit{motive : Ordinal → Sort*}(o : Ordinal)(H₁ : motive 0)(H₂ : (o : Ordinal) → motive o → motive (o + 1))(H₃ : (o : Ordinal) → Order.IsSuccLimit o → ((o' : Ordinal) → o' < o → motive o') → motive o)(h : Order.IsSuccLimit o) :Ordinal.limitRecOn o H₁ H₂ H₃ = H₃ o h fun x _h ↦ Ordinal.limitRecOn x H₁ H₂ H₃
Why only part: Lean's limit datum may also use the ordinal and the proof that it is a limit, and the earlier values reach it as a family indexed by the smaller ordinals.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 3 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Note: Lean lets the type of the object attached to an element depend on that element, the constant choice being the reading the passage describes.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit c9afcdd.
theorem WellFounded.fix_eq{α : Sort u}{C : α → Sort v}{r : α → α → Prop}(hwf : WellFounded r)(F : (x : α) → ((y : α) → r y x → C y) → C x)(x : α) :hwf.fix F x = F x fun y x ↦ hwf.fix F y
Note: It states the defining equation of the recursion at each element of the relation's domain.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 2 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit c9afcdd.
Why a special case: It is the case in which the well-founded relation is the strict order of an ordered type.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 3 for this passage, medium confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
More generally, one can define objects by transfinite recursion on any well-founded relation R. (R need not even be a set; it can be a proper class, provided it is a set-like relation; i.e. for any x, the collection of all y such that yRx is a set.)
Searched by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: the search receipt lists 4 queries and 3 candidates examined (receipt 3322430bae05).
Not re-checked: sampled out of challenge run chal-tierA-b1 by jack, in which Claude Opus 5 (claude-opus-5) re-checked a sample on 1 Nov 2026 (UTC).
Related, but states something else
WellFounded.fixrelated
def WellFounded.fix{α : Sort u}{C : α → Sort v}{r : α → α → Prop} :WellFounded r →((x : α) → ((y : α) → r y x → C y) → C x) →(x : α) →C x
How it differs: It is stated for a relation on a type, so neither the proper-class case nor the set-like condition the passage names has a counterpart in it.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, medium confidence.
Not re-checked: sampled out of challenge run chal-tierA-b1 by jack, in which Claude Opus 5 (claude-opus-5) re-checked a sample on 1 Nov 2026 (UTC).
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit c9afcdd.
Construction · supporting · Relationship to the axiom of choice
◔Partial
Partial: Mathlib states part of this (proved).
The following construction of the Vitali set shows one way that the axiom of choice can be used in a proof by transfinite induction: First, well-order the real numbers (this is where the axiom of choice enters via the well-ordering theorem), giving a sequence ⟨rα ∣ α < β⟩, where β is an ordinal with the cardinality of the continuum.
Why only part: The declaration is stated for an arbitrary type and produces the well-order together with the ordinal; naming that ordinal's cardinality as the continuum still needs the separate lemmas on the cardinality of an ordinal and on the cardinality of the reals.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
Why only part: This is the choice-powered well-ordering theorem for an arbitrary type, and it says nothing about the length of the resulting enumeration or about the continuum.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 2 for this passage, high confidence.
Not re-checked: sampled out of challenge run chal-tierA-b1 by jack, in which Claude Opus 5 (claude-opus-5) re-checked a sample on 1 Nov 2026 (UTC).
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
How it differs: It identifies the cardinality of the reals with the continuum, which is what fixes the length of the passage's enumeration, but on its own it asserts nothing about well-ordering or about indexing by ordinals.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 3 for this passage, high confidence.
Not re-checked: sampled out of challenge run chal-tierA-b1 by jack, in which Claude Opus 5 (claude-opus-5) re-checked a sample on 1 Nov 2026 (UTC).
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
How it differs: Taken in the reals with the rationals as subgroup, this proves that a set of coset representatives (a Vitali set) exists, built by picking one element from each coset. The passage claims something else: that the sequence from the article's transfinite recursion lists such a set. The declaration says nothing about that recursion or its output, so it is related but does not formalize the passage.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 1 for this passage, medium confidence.
Not re-checked: sampled out of challenge run chal-tierA-b1 by jack, in which Claude Opus 5 (claude-opus-5) re-checked a sample on 1 Nov 2026 (UTC).
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 3dfb8e2.
How it differs: It asserts that some subset of the reals is not Lebesgue measurable, which is the use the article's construction is usually put to, whereas the passage asserts the coset-representative property of the constructed set.
Details
Found by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: candidate 2 for this passage, high confidence.
Re-checked by Claude Opus 5 (claude-opus-5) on 1 Nov 2026 (UTC), challenge run chal-tierA-b1 by jack: it survived.
Record set tierA-b1b, exported 1 Nov 2026 by formal-companion d85385a.
Lean code: the declaration's type as Lean prints it, laid out as a header (the corpus keeps no hypothesis names), from the declaration corpus of universe u1 at commit 64c2773.
Searched by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: the search receipt lists 4 queries and 3 candidates examined (receipt 76cb3056f613).
Not re-checked: sampled out of challenge run chal-tierA-b1 by jack, in which Claude Opus 5 (claude-opus-5) re-checked a sample on 1 Nov 2026 (UTC).
Theorem · supporting · Relationship to the axiom of choice
✕Not formalized
Because there are models of Zermelo–Fraenkel set theory of interest to set theorists that satisfy the axiom of dependent choice but not the full axiom of choice, the knowledge that a particular proof only requires dependent choice can be useful.
Searched by Claude Opus 5 (claude-opus-5) on 31 Oct–1 Nov 2026 (UTC), locate run loc-tierA-b1b by jack: the search receipt lists 4 queries and 2 candidates examined (receipt 36f2f9de60a3).
Not re-checked: sampled out of challenge run chal-tierA-b1 by jack, in which Claude Opus 5 (claude-opus-5) re-checked a sample on 1 Nov 2026 (UTC).