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.

Coherent information equals output entropy minus environment entropy

Statement

Let N:A→B be a finite-dimensional channel, let V:A→B⊗E be a finite Stinespring isometry, let Nc(X)=Tr⁡B(VXV∗), and let ρ be a density operator on A. Then

Ic(ρ,N)=S(N(ρ))−S(Nc(ρ)).

The right-hand side is independent of the chosen Stinespring dilation, since it equals the purification-defined quantity in Complementary channels and coherent information.

Facts & Assumptions

Given: Nonzero finite-dimensional complex inner-product spaces A,B,E, a channel N:A→B, a Stinespring isometry V:A→B⊗E for N, and a density operator ρ on A.

[F1]

Complementary channels and coherent information: for a purification ∣ψ⟩∈A⊗R, Ic(ρ,N)=S(ωB)−S(ωBR), with ωBR=(N⊗id⁡R)(∣ψ⟩⟨ψ∣); the chosen complement is Nc(X)=Tr⁡B(VXV∗).

[F2]

Ensemble and purification and Purifications exist and are unitarily equivalent after padding: a purification is a unit vector ∣ψ⟩∈A⊗R with Tr⁡R∣ψ⟩⟨ψ∣=ρ, and such a finite purification exists.

[F3]

Schmidt decomposition: a finite pure bipartite vector has a Schmidt decomposition, so its two reduced density operators have the same positive eigenvalues with multiplicities.

[F4]

Every finite-dimensional channel has a finite isometric dilation: N(X)=Tr⁡E(VXV∗) for every input operator X.

[F5]

Finite-dimensional von Neumann entropy: S(τ)=−∑λj>0λjln⁡λj for the eigenvalues of a finite-dimensional density operator; zero eigenvalues contribute zero.

[F6]

Partial trace over a finite tensor factor, Partial trace over the first tensor factor, Ordered tensor registers, their inner product, and wire permutations, and Finite wire permutations are unitary basis permutations: ordered partial traces sum equal traced-factor indices; the factors of B⊗E⊗R can be reordered by a unitary wire permutation.

Proof

1.1givenF1F2F4F6construct

Choose a finite purification ∣ψ⟩∈A⊗R of ρ using [F2], and put ∣Ω⟩=(V⊗IR)∣ψ⟩∈B⊗E⊗R. Since V is an isometry, ∣Ω⟩ is a unit vector. Reorder to B⊗R⊗E using [F6]; tracing E gives ωBR because [F4] says Tr⁡E(VXV∗)=N(X). Tracing B and R gives Tr⁡B(VρV∗)=Nc(ρ): in the ordered product basis, the two finite partial-trace sums contract the R index to Tr⁡R∣ψ⟩⟨ψ∣=ρ and then contract the B index.

2.1F1F3F5step 1.1algebra∎

Regard the reordered ∣Ω⟩ as a bipartite vector across (B⊗R)∣E. By [F3], the positive eigenvalues of its BR and E marginals coincide. The entropy formula [F5] therefore gives S(ωBR)=S(Nc(ρ)). Substitute this identity into [F1] to obtain the displayed formula. As the left side of the formula is independent of both the purification and the chosen complement, the right side has the same value for every finite dilation.

Remarks

The equality is a consequence of the purity of the joint reference-output-environment state. It does not require a minimal Stinespring environment.

Depends on

Used by

Dependency tree · two levels

55 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