Alphabeta Math
TheoremStatement: AI-adaptedProof: AI-generatedPipeline-generatedprecheck passverified 2026-10-22 (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.

Form min-max below the essential spectrum

Statement

Assume the Axiom of Choice and Countable Choice. Let H≠{0} be a complex Hilbert space and let A be a self-adjoint operator with A≥−cI for some c≥0. Let EA be its spectral projection-valued measure, let Q(A)=D((A+cI)1/2), and define the associated form by qA[u]=∥(A+cI)1/2u∥2−c∥u∥2(u∈Q(A)). Set α=inf⁡σess(A), with α=+∞ when the essential spectrum is empty. For each integer k≥1, put λk(A)=inf⁡L⊆Q(A) linear, dim⁡L=k sup⁡u∈L∥u∥=1qA[u], where the infimum of an empty family is +∞. Then λk(A) is the kth eigenvalue below α, counted with multiplicity, when at least k such eigenvalues exist; otherwise λk(A)=α. If the inclusion Q(A)↪H is compact in the form norm ∥u∥Q(A)2=qA[u]+(c+1)∥u∥2, then A has a complete orthonormal eigenbasis; every eigenvalue has finite multiplicity, and the only possible accumulation is at +∞.

Facts & Assumptions

Given: Choice, a nonzero complex Hilbert space H, a self-adjoint A≥−cI, its spectral PVM EA, and the form domain Q(A).

[F1]

The associated form is closed, has domain D((A+cI)1/2), and satisfies qA[u]=∫(λ+c) dEu(λ)−c∥u∥2; its form norm with one additional unit shift is ∥(A+cI+I)1/2u∥ (Semibounded closed-form representation and perturbation, Unbounded Borel functional calculus: domains, products, spectral mapping).

[F2]

The discrete spectrum consists of isolated eigenvalues of finite multiplicity; the essential spectrum is its complement. For each real s, λ∈σess(A) exactly when every interval about λ has infinite-rank spectral projection, and σess(A) is closed (Discrete and essential spectrum of a self-adjoint operator).

[F3]

The spectral PVM is carried by σ(A), and its increasing interval projections converge strongly to I; bounded real Borel functions of A obey the product and norm identities. The spectral-calculus characterization of σ(A) says that t∈σ(A) exactly when every open interval about t has nonzero spectral projection, so σ(A) is closed (Spectral theorem for unbounded self-adjoint operators (PVM form), Unbounded Borel functional calculus: domains, products, spectral mapping).

[F4]

A positive compact self-adjoint operator has finite-dimensional nonzero eigenspaces, only finitely many eigenvalues outside any neighborhood of zero, and the closed span of its nonzero eigenspaces is the orthogonal complement of its kernel (Spectral theorem for compact self adjoint operators).

[F5]

A compact inclusion followed by a bounded operator is compact; compact resolvents are bounded compact operators on H (Compact linear operator, Compositions with a compact operator are compact).

Proof

Given: [A1] and the operator, PVM, and form in the Statement.

1.1A1F1F3given

By Choice [A1], [F1] and the spectral calculus give, for u∈Q(A), qA[u]=∫λ dEu(λ); the lower bound ensures this integral is finite after the stated shift. The form norm in the Statement is exactly ∥(A+(c+1)I)1/2u∥.

1.2F2F3givenalgebra

Fix t<α. Since A≥−cI, the spectral points in (−∞,t] lie in the compact interval [−c,t]. If this interval contained infinitely many spectral points, compactness of the interval and closedness of σ(A) would give an accumulation point s≤t in σ(A). Such an accumulation point is essential by [F2], contradicting s≤t<α. Thus there are finitely many spectral points below t, each isolated with finite multiplicity. By [F3], EA((−∞,t)) is supported on their eigenspaces, so it has finite rank.

1.3F3F4F5givenalgebra

Now assume the form-domain inclusion is compact. The map (A+(c+1)I)−1:H→Q(A) is bounded because ∥(A+(c+1)I)1/2(A+(c+1)I)−1f∥≤∥f∥ by [F3]. Composing with the compact inclusion Q(A)↪H shows that this resolvent is compact by [F5]. It is positive, self-adjoint and injective. By [F4], its nonzero eigenspaces have finite dimension and span H. On each such subspace, A=(A+(c+1)I)−(c+1)I has eigenvalue r−1−(c+1) when the resolvent eigenvalue is r>0. The compact spectral theorem says these r can accumulate only at zero; therefore the resulting eigenvalues of A have finite multiplicity and can accumulate only at +∞. This proves the compact-embedding conclusion.

2.1F2F3step 1.2algebra

Define νk=inf⁡{t∈R:rank⁡EA((−∞,t))≥k}, with inf⁡∅=+∞. If there are at least k eigenvalues below α, their finite-rank spectral projections show that νk is the kth such eigenvalue counted with multiplicity. If fewer exist, no t<α has rank at least k. When α<∞, closedness of σess(A) gives α∈σess(A), so every t>α contains an interval about α inside (−∞,t) whose spectral projection has infinite rank by [F2]; hence νk=α. When α=+∞, the spectrum consists only of discrete eigenvalues; if fewer than k exist, their finite-dimensional eigenspaces span H by [F3], so dim⁡H<k, no k-dimensional subspace exists, and νk=+∞. Thus νk has exactly the target value in every case.

3.1F1step 2.1algebra

Let L⊂Q(A) have dimension k, and set mL=sup⁡u∈L, ∥u∥=1qA[u]. By finite-dimensional Gram--Schmidt, choose an orthonormal basis e1,…,ek of L. For a unit vector u=∑iaiei, each ∣ai∣≤1 and qA[u]=∑i,jaiaj‾qA(ei,ej), so ∣qA[u]∣≤∑i,j∣qA(ei,ej)∣<∞ uniformly on the unit sphere. Thus mL<∞. For any t>mL, if rank⁡EA((−∞,t))<k, the projection restricted to L has a nonzero kernel; choose a unit u∈L with EA((−∞,t))u=0. Its spectral measure is supported in [t,∞), so [F1] gives qA[u]≥t>mL, a contradiction. Hence rank⁡EA((−∞,t))≥k, so νk≤t for every t>mL and therefore νk≤mL. Taking the infimum over L yields νk≤λk(A).

4.1F1F3step 2.1step 3.1algebra

If νk<∞ and t>νk, monotonicity of the interval projections gives some s<t with rank⁡EA((−∞,s))≥k. Choose a k-dimensional subspace L of this range. Its spectral support is bounded below by −c and above by s, hence L⊂Q(A); [F1] also gives qA[u]≤s∥u∥2 on L. Thus λk(A)≤s<t; letting t decrease to νk gives λk(A)≤νk. If νk=+∞, Step 2.1 shows dim⁡H<k, so the variational infimum is also +∞. Combining with Step 3.1 gives λk(A)=νk, proving the min-max identity and its spectral interpretation.

5.1step 1.1step 1.2step 2.1step 3.1step 4.1step 1.3∎

Steps 1.1-4.1 establish the form min-max characterization, including the essential-spectrum threshold and the empty-spectrum convention; Step 1.3 proves the complete discrete eigenbasis under compact form-domain embedding.

Depends on

Dependency tree · 2 levels

Sources