Alphabeta Math
LemmaStatement: Literature-sourcedProof: AI-adaptedPipeline-generatedjudge pass (gpt-6-luna)
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.

Partial trace contracts the trace norm on Hermitian operators

Statement

Let HA,HE be finite-dimensional complex inner-product spaces and let X=X† be an endomorphism of HA⊗HE. Then ∥Tr⁡EX∥1≤∥X∥1. The trace norm and partial trace are those of Trace Norm and Partial trace over a finite tensor factor. If either factor is zero-dimensional, both sides are zero.

Facts & Assumptions

Given: Finite-dimensional complex-inner-product spaces HA,HE and a Hermitian endomorphism X of HA⊗HE.

[F1]

Trace Norm: ∥Z∥1 is the sum of the singular values of Z.

[F2]

Trace Norm Spectral Formula: ∥Z∥1=tr⁡Z†Z.

[F4]

Partial trace over a finite tensor factor: in ordered orthonormal product bases, (Tr⁡EX)ii′=∑jX(i,j),(i′,j).

[F5]

The partial trace is characterized by the trace pairing and is basis-independent: tr⁡((Tr⁡EX)Y)=tr⁡(X(Y⊗IE)) for every endomorphism Y of HA.

[F6]

Tensor products of finitely many linear maps: (Y⊗IE)(x⊗e)=Yx⊗e on elementary tensors.

[F7]

Ordered tensor registers, their inner product, and wire permutations: the product inner product and ordered product basis are the finite tensor-product convention.

[F8]

The operator norm is zero on the zero domain and otherwise is max_{||v||=1} ||Tv||: ∥Y∥=max⁡∥x∥=1∥Yx∥ on a nonzero domain.

[F10]

Every finite-dimensional real or complex inner product space has an orthonormal basis: each finite-dimensional inner-product factor has an orthonormal basis.

[F12]

The basis-independent trace of an endomorphism of a finite-dimensional vector space: trace is the sum of diagonal entries in any ordered basis.

Proof

Given: As in Facts & Assumptions.

1.1F1F2F3F4F8F9F11F12choosecasesalgebra

If either tensor factor is zero-dimensional, then X=0 and the claimed inequality follows from the empty coordinate sum in [F4]; assume both factors are nonzero for the rest of the proof. For any Hermitian Z on a nonzero finite-dimensional complex-inner-product space, [F3] gives an orthonormal eigenbasis (ei) with eigenvalues λi. The adjoint identity [F11] and conjugate symmetry make λi=⟨Zei,ei⟩=⟨ei,Zei⟩=λi‾, so each λi is real. For every Hermitian Y with ∥Y∥≤1, [F12] gives tr⁡(ZY)=∑iλi⟨Yei,ei⟩root. Again [F11] makes each diagonal entry real, and [F8]–[F9] put it in [−1,1]. Thus ∣tr⁡(ZY)∣≤∑i∣λi∣=∥Z∥1: in the eigenbasis, Z†Z has eigenvalues λi2, so [F2] its nonnegative square root has eigenvalues ∣λi∣, and [F1], [F2], [F12] give the trace norm sum. Equality is attained by the Hermitian contraction acting by sgn⁡(λi) on each ei (and by 0 when λi=0). Hence ∥Z∥1=sup⁡Y=Y†, ∥Y∥≤1∣tr⁡(ZY)∣.

1.2givenF4F6F7F8F10F11algebra

Formula [F4] shows Tr⁡EX is Hermitian: transposing and conjugating its (i,i′) entry gives ∑jX(i′,j),(i,j)‾=∑jX(i,j),(i′,j). If Y=Y† and ∥Y∥≤1, choose an orthonormal basis (fj) of HE by [F10] and write arbitrary ξ=∑jxj⊗fj and ζ=∑jyj⊗fj. The product inner product and tensor-map action [F6]–[F7] give ⟨(Y⊗IE)ξ,ζ⟩=∑j⟨Yxj,yj⟩=∑j⟨xj,Yyj⟩=⟨ξ,(Y⊗IE)ζ⟩ by self-adjointness and [F11], so Y⊗IE is Hermitian. Also [F6]–[F8] give ∥(Y⊗IE)ξ∥2=∑j∥Yxj∥2≤∑j∥xj∥2=∥ξ∥2, so ∥Y⊗IE∥≤1.

2.1F1F4F5step 1.1step 1.2∎

For the remaining nonzero-factor case, apply the variational identity in step 1.1 to Tr⁡EX. For each allowed Y, [F5] and step 1.2 give ∣tr⁡((Tr⁡EX)Y)∣=∣tr⁡(X(Y⊗IE))∣≤∥X∥1, using step 1.1 for the Hermitian contraction Y⊗IE. Taking the supremum proves the claimed inequality.

Depends on

Used by

Dependency tree · two levels

48 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