Section V remark after Lemma 2: zero-determinant factors
OpenQAlgorithms.Papers.Independent.Support50.P28.singular_factorsConsider finitely many complete bosonic modes, with no photon cutoff or energy constraint. Covariances use the interleaved coordinates , , and doubled vacuum covariance . A centered Gaussian channel is an actual completely positive trace-preserving map, with covariance action and , where are real, , and .
Write for the normalized centered Gaussian densities and, for real , put . For a finite channel family, multiplicativity means equality of the joint Gaussian-input supremum to the product of the factor suprema; joint inputs may be entangled.
Let and let be single-mode Gaussian channels with for every . For every real , their tensor product is multiplicative. Moreover, if another physical single-mode family has the same noise matrices and for every , then
Thus the supremum in this regime depends only on the noise, not on the zero-determinant matrices . Infinite squeezing may approximate the supremum without attaining it.
import Definitions.Def_QAlgorithms_Papers_Independent_Support50_P28
noncomputable section
open scoped BigOperators ENNReal ComplexOrder Matrix
namespace QAlgorithms.Papers.Independent.Support50.P28
attribute [local instance] Classical.propDecidable Classical.decEq
open QAlgorithms.RealRegisters (Hilbert)
open QAlgorithms.Papers.Independent.Dependency50.P33
(State IsDensity IsTraceClassChannel rankOne)
open QAlgorithms.Papers.Independent.Claimed50.P47
(Occupation gaussianStateCharacteristic singleCharacteristic weylKernel)
open QAlgorithms.Papers.Independent.Claimed50.P30 (phiMinus)
/-- PDF p.5, remark after Lemma 2: all-zero determinant factors are multiplicative
and the supremum does not depend on X. Comparison to every physical family with
the same Y expresses independence without assuming attainment. -/
theorem singular_factors {r : ℕ} (hr : 0 < r)
(E : Fin r → CenteredOscillatorChannel (Fin 1)) (hzero : ∀ i, (E i).X.det = 0)
(p : ℝ) (hp : 1 < p) :
GaussianPurityMultiplicative p E ∧
∀ F : Fin r → CenteredOscillatorChannel (Fin 1),
(∀ i, (F i).X.det = 0 ∧ (F i).Y = (E i).Y) →
maximalOutputPurity p (tensorOscillatorChannel E) = maximalOutputPurity p (tensorOscillatorChannel F) := by sorry
end QAlgorithms.Papers.Independent.Support50.P28
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
Let , let be real, and let , , be centered one-mode oscillator channels with for every . The terms in this assertion have the following meanings. A nuclear coefficient matrix on admits with and ; a state is such a matrix with real nonnegative finite quadratic forms and summable real diagonal of sum . For a finite mode set , use , the block matrix , and
Write for states admitting a real symmetric with , summable inner and outer characteristic series for every , and . A centered oscillator channel has real matrices satisfying and ; its map preserves nuclear matrices and is additive and complex homogeneous on them, and applying to each block of any state with a finite ancilla gives a state. All finite ancilla sizes, including the vacuous size , are included. Its specified output state has coefficients , summable characteristic series, and
for every input state. Let be a chosen bounded operator with coordinate coefficients when one exists, and otherwise, and set
where operator powers use continuous functional calculus and in . On the combined mode set , define by applying the maps successively in increasing block order, with
Then
Moreover, for every other family of centered one-mode channels , with maps , if and for every , then
The hypotheses permit arbitrary singular , including the zero matrix; they specify equality of the noise matrices themselves for the second conclusion.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.