Theorem 5: EPR–GHZ region for isometric broadcast channels
OpenQAlgorithms.Papers.Independent.Gap48.P39.epr_ghz_capacityLet be a rectangular isometry on positive-dimensional finite systems. Define as the closed unassisted vanishing-error region with coordinate order : tripartite Alice–Bob–Charlie GHZ ebits and independent Alice–Bob EPR ebits. A block code has one unit pure encoding and separate local CPTP decoders, applied after ; one joint squared-fidelity test compares the decoded state with the product of GHZ rank and EPR rank . For the one-letter pure-ensemble output , define the nonnegative information region by
This region is contained in , and imposing does not change it. The closure of its positive-blocklength rescaled union equals . Entropies are in bits. This establishes the full regularized EPR–GHZ region with a joint product target, independently of two-EPR transmission.
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]
/-- Theorem 5, PDF pp.11–12,14. Isometric U; rate order (GHZ,EPR_B), one
joint product-target fidelity, one-letter state, cap, and closed regularization. -/
theorem epr_ghz_capacity [Nonempty A] [Nonempty B] [Nonempty C]
(U : Matrix (B × C) A ℂ) (hU : Uᴴ * U = 1) :
quantumInformationRegion (conjSuperop U) false ⊆ qqCapacity (conjSuperop U) ∧
quantumInformationRegion (conjSuperop U) false = quantumInformationRegion (conjSuperop U) true ∧
qqCapacity (conjSuperop U) = closure (regularizedQuantumRegion (conjSuperop U)) := by sorry
end QAlgorithms.Papers.Independent.Gap48.P39
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
Let be nonempty finite index types with decidable equality, and let satisfy . Set . For an arbitrary matrix , write , where is applied by continuous functional calculus. Define , for , as follows: a pair belongs to it precisely when both coordinates are nonnegative and there exist positive integers , weights summing to , and unit vectors , such that, on putting , , , and , one has
When , additionally require ; when , this additional requirement is absent. Define to be the closure of the nonnegative pairs with the following property: for every , there is such that for every , with and , there exist a unit vector on , a channel from to , and a channel from to , whose output after , leaving the first two reference registers untouched, has squared fidelity at least with an EPR state of rank between the first reference register and Bob’s first output, tensored with the pure GHZ state on the second reference register and the remaining two outputs. Here a channel has a finite Kraus representation with ; squared fidelity is ; and the EPR vector of rank is . The assertion is
All closures are in the usual topology on ; zero rates are included, and give rank in the corresponding code register.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.