Alphabeta Math
LemmaStatement: Literature-sourcedProof: AI-adaptedPipeline-generatedprecheck pending
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.

Entropy bounds for a finite mixture of states

Statement

For states ρi and finite probabilities pi, put ρˉ=∑ipiρi. Then 0≤S(ρˉ)−∑ipiS(ρi)≤H(p).

Facts & Assumptions

Given: A common nonzero finite-dimensional space H, a finite probability vector (pi), and density operators ρi on H. The Shannon entropy is in nats: H(p):=−∑i:pi>0piln⁡pi, with 0ln⁡0:=0.

[F1]

Finite-dimensional von Neumann entropy: Finite-dimensional von Neumann entropy.

[F2]

Nonnegativity and equality for finite quantum relative entropy: Nonnegativity and equality for finite quantum relative entropy.

[F3]

Schmidt decomposition: Schmidt decomposition.

[F4]

The basis-independent trace of an endomorphism of a finite-dimensional vector space: 'The basis-independent trace of an endomorphism of a finite-dimensional vector space'.

[F6]

Finite-dimensional quantum relative entropy and Order, continuity, range, and the product, quotient, and reciprocal laws for the natural logarithm: on support inclusion, D(α∥β)=tr⁡(αln⁡α)−tr⁡(αln⁡β), and ln⁡(xy)=ln⁡x+ln⁡y for positive x,y.

Proof

1.1F1F2F4F5F6givenalgebraconstruct

Omit zero weights. For each i with pi>0, ρˉ−piρi=∑j≠ipjρj is non-negative. If v∈ker⁡ρˉ, then 0=⟨ρˉv,v⟩=pi⟨ρiv,v⟩+∑j≠ipj⟨ρjv,v⟩. All terms are nonnegative, so pi>0 implies ⟨ρiv,v⟩=0 and [F5] gives ρiv=0. Hence supp⁡ρi⊆supp⁡ρˉ and every D(ρi∥ρˉ) is finite. The trace formula [F6], linearity of trace [F4], and ρˉ=∑ipiρi give ∑ipiD(ρi∥ρˉ)=S(ρˉ)−∑ipiS(ρi)≥0 by [F2].

2.1F1F2F3F4F6givenalgebraconstructstep 1.1

For the upper bound, diagonalize each ρi=∑jλij∣vij⟩⟨vij∣, omitting zero weights and eigenvalues, and purify the mixture as ∣v⟩=∑ijpiλij∣vij⟩⊗∣i,j⟩. Its label marginal G is a state with diagonal qij=piλij and has the same nonzero eigenvalues as ρˉ by Schmidt decomposition. The logarithm product law in [F6] gives H(q)=H(p)+∑ipiS(ρi).

3.1F1F2F3F4F5F6givenalgebraconstructstep 2.1∎

For Δ=diag⁡(q), its kernel is spanned by coordinate labels with qij=0. Each such basis vector eij has ⟨Geij,eij⟩=qij=0, so [F5] gives Geij=0. Hence G annihilates ker⁡Δ and supp⁡G⊆supp⁡Δ. The formula [F6] now gives 0≤D(G∥Δ)=H(q)−S(G)=H(p)+∑ipiS(ρi)−S(ρˉ). This proves both bounds, including singleton/zero-weight cases.

Depends on

Used by

Dependency tree · two levels

37 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