The Holevo difference is independent of the pure decomposition
OpenQAlgorithms.Papers.Independent.Gap48.P43.pureDecomposition_holevoDifference_independentFix a unitary interaction of a quantum register with one complete environment in a normalized pure initial state. Let and be finite or countable probability ensembles of normalized input states, pure whenever their probability is nonzero. Their computed average inputs are equal:
For each ensemble and each output , compute the conditional and average output states from that fixed interaction and put , with entropies in bits. Assume the average output entropies and the weighted conditional entropy sums appearing in these differences are finite. Then
Thus the difference depends only on the average density operator. The assertion permits different alphabets and probabilities, repeated and nonorthogonal signals, and channels for which one or both individual Holevo quantities are constant.
import Definitions.Def_QAlgorithms_Papers_Independent_Gap48_P43
noncomputable section
open MeasureTheory
open scoped ENNReal BigOperators
namespace QAlgorithms.Papers.Independent.Gap48.P43
open QAlgorithms.RealRegisters (Hilbert)
open QAlgorithms.Papers.Independent.Dependency50.P33
(State PureState productCoeff)
open QAlgorithms.Papers.Independent.Frontier18.P01
(CountableStateEnsemble boundedConjugation traceSecond traceFirst)
open QAlgorithms.Papers.Independent.Frontier18.P18
(boundedOperator operatorCoeff)
open QAlgorithms.Papers.Independent.Claimed50.P35 (spectralEntropyBits)
open QAlgorithms.Papers.Independent.Frontier30.P10 (HilbertPOVM)
attribute [local instance] Classical.propDecidable Classical.decEq
/-- Schumacher--Westmoreland, quant-ph/9709058v2, PDF p.5, paragraph after (12).
Two pure discrete decompositions with the same computed average have the same
Holevo difference. No nonconstant dependence of either individual chi is asserted. -/
theorem pureDecomposition_holevoDifference_independent {Q E : Type}
(d : PureDilation Q E) (a b : CountableStateEnsemble Q)
(ha : ActivePure a) (hb : ActivePure b)
(haverage : averageInput a = averageInput b)
(haBob : FiniteOutputEntropies a (bobOutput d))
(haEve : FiniteOutputEntropies a (eveOutput d))
(hbBob : FiniteOutputEntropies b (bobOutput d))
(hbEve : FiniteOutputEntropies b (eveOutput d)) :
holevoBits a (bobOutput d) - holevoBits a (eveOutput d) =
holevoBits b (bobOutput d) - holevoBits b (eveOutput d) := by sorry
end QAlgorithms.Papers.Independent.Gap48.P43
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
Let be arbitrary types and . A state on means a matrix admitting with and summable , having real nonnegative quadratic forms for every finite and every , and having a summable real diagonal with sum . Fix consisting of a state on , expressible as for a unit vector , and a surjective complex linear isometry . With , define
Let and be ensembles of states on , with extended nonnegative weights satisfying . Assume every with , and every with , has coefficients for some unit vector ; zero-weight states have no additional purity requirement. Define and , where converts extended nonnegative reals to reals, sending to , and assume the matrices and are equal. For any matrix on , choose a bounded complex linear operator with coefficients when one exists, and otherwise use zero, and define
using the complex continuous functional calculus and extended nonnegative clipping . Set
Assume all four finiteness conditions: for each ensemble and each output map , its average-output entropy and its weighted sum of individual output entropies are both strictly below . Then
No matching of the individual states or weights of the two ensembles is required, and no countability or finite-dimensionality assumptions are imposed on . Real and complex infinite sums are totalized to zero when nonsummable.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.