Alphabeta Math
TheoremStatement: Literature-sourcedProof: AI-adaptedPipeline-generatedprecheck passjudge pass (gpt-6.1-sol)
How statement and proof provenance work

The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.

  • Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
  • AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
  • AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.

These labels describe origin, not correctness: citations and verification chips remain separate evidence.

Primitive tableau idempotents by Jucys-Murphy interpolation

Statement

Let T be a standard tableau of size n, let μ be the shape of the restriction of T to [n−1], and let A(μ) be the set of contents of the addable nodes of μ. Set P[1]:=1 and define recursively, for n≥2, PT:=PT↓[n−1]∏c∈A(μ)c≠cT(n)Xn−ccT(n)−c∈C[Sn]. Then all displayed denominators are nonzero, PT is the rank-one idempotent projecting onto the Young line CvT, each PT is a polynomial in X1,…,Xn with rational coefficients, and the PT over the standard tableaux of size n are pairwise orthogonal idempotents with ∑TPT=1.

Facts & Assumptions

Given: The chain S1⊂⋯⊂Sn, the Jucys-Murphy elements Xk, and, for every standard tableau T of size n, the Young line CvT=im⁡PT with the idempotents PT of the diagonal-algebra theorem (The Gelfand-Tsetlin algebra is the diagonal algebra of the Young basis, The Jucys-Murphy elements of the symmetric group algebra).

[F1]

GZ(n)=⨁TCPT over the standard tableaux T of size n; the PT are nonzero pairwise orthogonal idempotents with ∑TPT=1, and CvT=im⁡PT is one-dimensional (The Gelfand-Tsetlin algebra is the diagonal algebra of the Young basis).

[F2]

XkvT=cT(k)vT for every standard tableau T and every k, with cT(k) the content of the node carrying k (The joint spectrum of the Jucys-Murphy elements is the set of tableau content vectors, The content of a node and the content vector of a standard tableau).

[F3]

For a partition μ, the contents of distinct addable nodes are distinct, and the addable nodes of μ are the nodes (i,μi+1) with i=1 or μi−1>μi, together with the new-row node (Distinct addable nodes of a partition have distinct contents, Removable and addable nodes).

[F4]

The restriction of T to [n−1] is a standard tableau of shape μ that is obtained by deleting the node carrying n; conversely every standard tableau S of size n with S↓[n−1]=T↓[n−1] is obtained by placing n in an addable node of μ (Tableaux and standard tableaux, Removable and addable nodes).

Proof

technique · induction
1.1F1givenbase

Base case. For n=1 the only standard tableau is [1], the line is Cv[1]=C⋅1, and P[1]:=1 is the rank-one projection onto it with rational (indeed integer) coefficients.

1.2F1givenih

Induction hypothesis. For every standard tableau S of size n−1 the recursively defined element PS∈C[Sn−1] equals the idempotent PS of [F1], is the projection onto CvS, and is a polynomial in X1,…,Xn−1 with rational coefficients.

1.3F3F4givenalgebra

Let T be a standard tableau of size n and μ:=shape⁡(T↓[n−1]). The set A(μ) of contents of addable nodes is finite and cT(n)∈A(μ): the node carrying n is an addable node of μ by [F4]. By [F3] the denominators cT(n)−c, c∈A(μ), c≠cT(n), are nonzero integers, so the displayed product is a well-defined element of C[Sn].

2.1step 1.2F1algebra

If S is a standard tableau of size n−1 with S≠T↓[n−1], then the factor PT↓[n−1] acts as 0 on the line CvS by step 1.2; hence PT acts as 0 on every Young line of size n whose restriction to [n−1] differs from T↓[n−1].

2.2step 1.2F2F4algebra

Now let S be a standard tableau of size n with S↓[n−1]=T↓[n−1], and let x be the addable node of μ carrying n in S, so that cS(n)=c(x) by [F4] and [F2]. Then PT↓[n−1] acts as the identity on CvS by step 1.2, and the interpolation factor acts on CvS by the scalar ∏c∈A(μ)c≠cT(n)cS(n)−ccT(n)−c, because Xn acts on CvS by cS(n) by [F2].

2.3step 1.2step 1.3algebra

Polynomial form. By step 1.2 the factor PT↓[n−1] is a polynomial in X1,…,Xn−1 with rational coefficients; each factor (Xn−c)/(cT(n)−c) is a polynomial in Xn with rational coefficients because cT(n)−c∈Z∖{0} by step 1.3; hence PT is a polynomial in X1,…,Xn with rational coefficients.

3.1F3step 1.3step 2.2algebra

The scalar of step 2.2 equals 1 when S=T, and equals 0 when S≠T: if S=T then cS(n)=cT(n) and every factor is (cT(n)−c)/(cT(n)−c)=1; if S≠T then S places n in a different addable node of μ, so cS(n)≠cT(n) by [F3], and the factor with c=cS(n) has numerator 0 while the denominator is nonzero by step 1.3.

4.1step 2.1step 3.1F1algebra

Steps 2.1-2.3 show that PT acts as the identity on the line CvT and as 0 on every other Young line of size n. Since GZ(n)=⨁SCPS is the algebra of operators diagonal in the Young basis by [F1], the element PT equals the idempotent PT of [F1], namely the rank-one projection onto CvT. In particular PT2=PT and PT≠0.

5.1step 4.1F1

Orthogonality and the partition of unity are inherited from [F1]: for T≠T′ the idempotents PT,PT′ of [F1] multiply to 0, and ∑TPT=1 over the standard tableaux of size n.

6.1step 1.1step 4.1step 5.1step 2.3discharge-induction∎

Steps 1.1, 2.1-2.3, 3.1 and 4.1 are the base, the successor and the conclusions of an induction on n; therefore the recursive formula defines the tableau idempotents for every n, with all properties asserted.

Remarks

  • Interpolation at the spectrum. The factor is the Lagrange polynomial that takes the value 1 at the content cT(n) and vanishes at the other contents in A(μ); the contents in A(μ) are pairwise distinct by [F3], and they are the eigenvalues of Xn on the lines over T↓[n−1] by [F2]. This is Garsia's recursion for the seminormal units, stated here for the tableau idempotents of the diagonal algebra.

  • Integrality fails only at the denominators. Over Z the formula must be cleared of denominators; over C the rational coefficients are harmless. The first nontrivial denominator occurs for n=2: the addable contents of (1) are 1,−1, giving the projectors (1+X2)/2 and (1−X2)/2.

  • Choice. The recursion selects no object: the addable node carrying n in a given tableau is determined by that tableau, and the idempotents are built from the fixed chain.

Depends on

Used by

Dependency tree · two levels

21 results within two dependency steps of this one, each drawn at its shortest distance from it. An arrow runs from a result to what uses it, so the chart reads left to right and ends at this result, which carries a heavier outline. Every node is a link to that result. Click elsewhere on the chart to enlarge it.

Sources