Theorem A.23: H(psi|phi)≤H(psi|phi∘phiPrime), in particular ≤H(psi), for compatible phi,psi and arbitrary CP-unital phiPrime:Xprime→X.
OpenQAlgorithms.Papers.Independent.Renew46.P43.knowledgeDecreasesUncertaintyFix a normalized state and compatible CP-unital Heisenberg maps , , together with any CP-unital map . Conditional entropy is joint viewed entropy minus conditioning entropy and need not be nonnegative.
This is the source result with its stated operational or information-theoretic domain.
import Definitions.Def_QAlgorithms_Papers_Independent_Renew46_P43
noncomputable section
namespace QAlgorithms.Papers.Independent.Renew46.P43
open scoped BigOperators Matrix ComplexConjugate ComplexOrder ENNReal Matrix.Norms.L2Operator
open Filter MeasureTheory
open QAlgorithms.Papers.Independent.Information100.P08 (IsPure)
open QAlgorithms.Papers.Independent.Information100.P06 (Below)
open QAlgorithms.Papers.Independent.Information100.P60 (copies relativeBits)
open QAlgorithms.Papers.Independent.Dependency50.P19 (count both)
open QAlgorithms.Papers.Independent.Information100.P22 (coherentInformation)
attribute [local instance] Classical.propDecidable Classical.decEq
open QAlgorithms.Papers.Independent.Claimed50.P40 (InBlocks)
attribute [local instance] QAlgorithms.Papers.Independent.Renew46.P43.matrixMeasurableSpace
/-- Theorem A.23, PDF p.75: conditioning is weakened by Heisenberg preprocessing,
with the unconditional upper bound retained as a second conclusion. -/
theorem knowledgeDecreasesUncertainty {X Y Xnew A : IntrinsicSystem}
(E : IntrinsicOperation X A) (F : IntrinsicOperation Y A) (hcompat : CommutingImages E.map F.map)
(i : IntrinsicOperation Xnew X) (rho : Matrix (Fin A.dim) (Fin A.dim) ℂ) (hrho : A.IsState rho) :
intrinsicConditionalEntropy Y X F.map E.map rho ≤
intrinsicConditionalEntropy Y Xnew F.map (E.map.comp i.map) rho ∧
intrinsicConditionalEntropy Y X F.map E.map rho ≤ viewedEntropy Y F.map rho := by sorry
end QAlgorithms.Papers.Independent.Renew46.P43
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
Let be systems, each with a positive integer dimension and a block-label function ; call matrices block diagonal when entries between different block labels vanish. Let , , and be complex-linear maps that preserve identity matrices and block diagonality. Each must also preserve positive semidefiniteness under every amplification, including amplification size : whenever a positive semidefinite block matrix has entries zero between different source-system block labels, applying the map separately to every matrix block of yields a positive semidefinite matrix. Assume for every block-diagonal for and for . Let be positive semidefinite, have trace , and be block diagonal for . For a map from the matrices of a system into those of , define its restricted matrix by
For maps associated with systems , respectively, define
where the are matrix units in their respective spaces. With , where is applied by functional calculus, both inequalities hold:
No injectivity, multiplicativity, or conjugate-transpose-preservation assumption is imposed on these maps beyond the stated requirements; dimensions equal to are included.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.