Lemma 5: product-target fidelity from a pure marginal
ProvedQAlgorithms.Papers.Independent.Gap48.P39.product_target_fidelityLet be a normalized density matrix on positive-dimensional finite systems, let be a unit vector, and let be any normalized density matrix. With the actual partial traces,
Here , is squared Uhlmann fidelity, and is the full trace norm. The lower bound can be negative and is not clamped. This controls joint fidelity to a product target from a pure marginal and the other marginal's trace-norm error.
import Definitions.Def_QAlgorithms_Papers_Independent_Gap48_P39
noncomputable section
open scoped BigOperators Matrix ComplexConjugate
namespace QAlgorithms.Papers.Independent.Gap48.P39
open QAlgorithms.Papers.Independent.Information100.P21 (Op)
open QAlgorithms.Papers.Independent.Information100.P03 (PureEnsemble)
open QAlgorithms.Papers.Independent.Claimed50.P08 (cqBlocks holevoInfo)
open QAlgorithms.Papers.Independent.Dependency50.P19 (both epr mutualInformation marginalRA marginalRB)
open QAlgorithms.Papers.Independent.Information100.P58 (Distribution)
open QAlgorithms.Papers.Independent.Information100.P67 (squaredFidelity conditionalEntropy groupedChannelPower)
attribute [local instance] Classical.propDecidable Classical.decEq
variable {A B C X T E : Type}
variable [Fintype A] [DecidableEq A] [Fintype B] [DecidableEq B]
variable [Fintype C] [DecidableEq C] [Fintype X] [DecidableEq X]
variable [Fintype T] [DecidableEq T] [Fintype E] [DecidableEq E]
/-- Lemma 5, PDF pp.11–12: squared fidelity, no square root on the factor 3;
the lower bound may be negative and is not clamped. -/
theorem product_target_fidelity [Nonempty A] [Nonempty B]
(rho : Op (A × B)) (hr : IsDensityMatrix rho)
(psi : EuclideanSpace ℂ A) (hp : IsState psi) (sigma : Op B) (hs : IsDensityMatrix sigma) :
1 - 3 * (1 - squaredFidelity (pureDensity psi) (traceRight rho)) -
traceNorm (traceLeft rho - sigma) ≤
squaredFidelity rho (Matrix.kronecker (pureDensity psi) sigma) := by sorry
end QAlgorithms.Papers.Independent.Gap48.P39
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
For nonempty finite types with decidable equality, let be a positive semidefinite, trace-one complex matrix on , let have Euclidean norm , and let be a positive semidefinite, trace-one complex matrix on . Put , define
with matrix square roots interpreted through continuous functional calculus. Then
The trace-norm term is not divided by , and the left side is not replaced by its maximum with zero.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.