Corollary 6.2 / (84): separable side channels are useless
OpenQAlgorithms.Papers.Independent.Gap20.P07.separableSideChannel_uselessLet and be CPTP channels between nonzero complex Hilbert spaces. Suppose factors operationally as measurement followed by preparation: a POVM with an arbitrary measurable outcome space has its actual probability law on every input state, and the output is the integrable barycenter of normalized prepared states. Then
No finite or countable outcome restriction and no assumption of a product input is imposed. This is the separable-side-channel consequence of the forward-assistance theorem.
import Definitions.Def_QAlgorithms_Papers_Independent_Gap20_P07
noncomputable section
open Filter MeasureTheory
open scoped Topology ENNReal NNReal Matrix ComplexOrder ComplexConjugate
attribute [local instance] Classical.propDecidable
attribute [local instance] Classical.decEq
attribute [local instance] IsStarNormal.instContinuousFunctionalCalculus
namespace QAlgorithms.Papers.Independent.Gap20.P07
open QAlgorithms
open QAlgorithms.RealRegisters (Hilbert)
open QAlgorithms.Papers.Independent.Dependency50.P33
(State IsTraceClass IsTraceClassChannel traceSize rankOne tensorApply)
open QAlgorithms.Papers.Independent.Frontier30.P14
(Operation isometryCoeff conjugateKernel)
open QAlgorithms.Papers.Independent.Claimed50.P35
(stabilizedTraceClassNorm traceClassTensorMap traceClassChannelPower
transpose transposeBound TraceLinear tracePair MeasurePrepare)
open QAlgorithms.Papers.Independent.Gap48.P29 (fe FiniteCPMap)
/-- PDF p.23, Corollary 6.2, (84): operational measurement--preparation side
channels with arbitrary outcome spaces; no countable-decomposition assumption. -/
theorem separableSideChannel_useless {I J K L : Type}
[Nonempty I] [Nonempty J] [Nonempty K] [Nonempty L]
(T : Operation I J) (R : Operation K L) (hT : IsTraceClassChannel T)
(hR : IsTraceClassChannel R) (hseparable : MeasurePrepare R) :
qubitReferenceCapacity (traceClassTensorMap T R) = qubitReferenceCapacity T := by sorry
end QAlgorithms.Papers.Independent.Gap20.P07
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
Let be arbitrary nonempty types. Call a matrix trace class when it has an expansion with and summable, and let be the infimum in of the corresponding nonnegative expansion costs, with empty infimum . A density matrix is trace class, has real nonnegative finite principal quadratic forms, and has summable real diagonal with sum . Suppose from matrices on to matrices on , and from matrices on to matrices on , preserve trace class matrices, are additive and complex homogeneous on them, and preserve densities under every finite ancillary block amplification. Suppose also that has the following measurement and preparation representation: there exist a measurable outcome type , bounded complex linear effects on for every set , with and real nonnegative quadratic forms for measurable , finite measures for all satisfying for measurable , probability measures for every density matrix on , and output densities on , such that all coordinates of are measurable,
and every coordinate integrand in the latter equality is integrable; here is the th coordinate of . For a map , set , where amplification applies to each ancillary block, and let be the supremum in of strictly positive real such that every pair of positive-integer sequences with and extended-real satisfies
where range over trace class channels encoding the reference register into the input and decoding the output back to it, tensor powers apply the map recursively to each coordinate, and an empty rate supremum is . Define the tensor map explicitly by
Then . No countability assumption is imposed on any of the four register types or on the measurement outcome type.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.