Existence of an explicit incomparable Gaussian pair
OpenQAlgorithms.Papers.Independent.Gap48.P47.incommensuratePairFor real , let
be the two-mode normal covariance, in the order with and vacuum covariance . There exist actual normalized Gaussian oscillator states with finite second moments such that
LOG denotes products of independent local Gaussian channels implemented by finite Gaussian ancillas, local Gaussian unitaries and partial traces. The negative final correlation is part of the example.
This establishes non-totality of operational convertibility without asserting antisymmetry on raw states.
import Definitions.Def_QAlgorithms_Papers_Independent_Gap48_P47
noncomputable section
open scoped BigOperators ComplexOrder
namespace QAlgorithms.Papers.Independent.Gap48.P47
open QAlgorithms.Papers.Independent.Dependency50.P33
(State IsTraceClass IsTraceClassChannel)
open QAlgorithms.Papers.Independent.Claimed50.P19
(Cov FockIndex DensityKernel GaussianWithCovariance CharacteristicSummable
multimodeWeylTrace covarianceCharacteristicValue modeCovarianceBlock
standardFormCovariance)
open QAlgorithms.Papers.Independent.Renew46.P39 (AffineMultimodeGaussianUnitary)
open QAlgorithms.Papers.Independent.Frontier30.P08 (interleavedSymplecticForm)
open QAlgorithms.Papers.Independent.Dependency50.P10 (complexify)
open QAlgorithms.Papers.Independent.Frontier30.P14 (boundedUnitaryConjugation)
attribute [local instance] Classical.propDecidable Classical.decEq
/-- PDF p.4, explicit application: actual Gaussian states with the two
specified normal covariances exist and are incomparable in BOTH directions.
The negative fourth coordinate is retained; Proposition 2 is not extended. -/
theorem incommensuratePair :
∃ rho target : TwoModeGaussianDensity,
rho.covariance = standardFormCovariance 2 2 1 1 ∧
target.covariance = standardFormCovariance 2 2 1 (-(1 / 2)) ∧
¬ LOGReachable rho target ∧ ¬ LOGReachable target rho := by sorry
end QAlgorithms.Papers.Independent.Gap48.P47
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
There exist two two-mode Gaussian densities and whose covariance matrices, in the coordinate order , are respectively
and neither density can be transformed into the other by the pair of local operations described below. Here, for modes, the occupation indices are , and the Hilbert space is . A kernel on an index set is called trace class in this statement when there are sequences such that and . A density kernel additionally satisfies, for every finite and every , that is real and nonnegative, and its real diagonal is summable with . A -mode Gaussian density with covariance consists of such a density kernel , a real positive definite matrix with positive semidefinite, and the following further properties: is summable, and there is a real vector such that for every real ,
For each , the inner series in this expression must be summable, and the series of its values over must also be summable. Here is block diagonal with blocks , and
The excluded transformation from to means the existence of functions on one-mode kernels, each satisfying the following channel and dilation conditions, with
The channel conditions are that each function sends trace-class kernels to trace-class kernels, is additive on pairs of trace-class kernels, respects every complex scalar on trace-class kernels, and, for every , its entrywise extension to an ancillary index sends every density kernel on the combined indices to a density kernel; explicitly this extension has entries . The dilation condition for , and separately for , is that there exist , an -mode Gaussian density kernel , and a complex linear isometric bijection on , such that for every trace-class ,
The tensor kernel has entries , with the system occupation first. Moreover, must come with a real matrix and real displacement vector satisfying and, for every real ,
The two nonexistence assertions apply separately to and to . The ancillary mode count is allowed: its occupation set is a singleton and products over its modes are empty products. The finite ancillary test also includes , where there are no density kernels and that test is vacuous. Infinite sums in the kernel formulas are the total sums used by the definitions, taking value zero for a nonsummable real or complex family; the stated summability requirements are exactly those specified above.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.