§III.C, Eq.(30): optimized coherent information of all N→M cloners
OpenQAlgorithms.Papers.Independent.Support50.P01.cloner_singleUseLet be integers. Let be the specified optimal universal qubit cloner on , with receiver support and environment dimension . Its common dilation on normalized occupation kets is exactly Eqs.(21),(22). Write , for coherent information in bits, for its supremum over all density inputs, and for its regularization over all positive block lengths and arbitrary entangled block inputs. For every such pair,
The normalized maximally mixed input dominates every density input. This retains the all- single-use claim of Eq.(30), without asserting the general conjugate-degradability conjecture or an all-cloner exact capacity formula.
Formalization Note The printed Eqs.(23),(25),(29) omit the input factor ; the normalized output in this statement has unit trace. The source only supplies the concavity argument for its two established families, so the wider single-use claim has a disclosed proof gap.
import Definitions.Def_QAlgorithms_Papers_Independent_Support50_P01
noncomputable section
namespace QAlgorithms.Papers.Independent.Support50.P01
open scoped BigOperators Matrix ComplexConjugate ComplexOrder
open Filter
open QAlgorithms.Papers.Independent.Information100.P67
(receiverChannel complementaryChannel IsDegradable dilationPower maximumCoherentInformation)
open QAlgorithms.Papers.Independent.Information100.P22 (coherentInformation quantumCapacity)
open QAlgorithms.Papers.Independent.Information100.P08 (maximallyMixed)
open QAlgorithms.Papers.Independent.Dependency50.P41 (EntanglementBreaking)
open QAlgorithms.Papers.Independent.Dependency50.P40 (outputFirstChoi)
attribute [local instance] Classical.propDecidable
attribute [local instance] Classical.decEq
variable {a b e c : ℕ}
/-- PDF pp.10–11, Eq.(30): all-N,M SINGLE-USE optimized information.
The source supplies concavity only for its two proven families; its wider Eq.(30)
claim is retained without adding the general conjecture as a hypothesis.
Corrected normalization: the omitted 1/(N+1) in Eqs.(23),(25),(29) makes the
normalized environment output I/(M-N+1), not (N+1)I/(M-N+1). -/
theorem cloner_singleUse (N M : ℕ) (hN : 1≤N) (hNM : N≤M) :
IsDensityMatrix (maximallyMixed (N+1)) ∧
receiverChannel (clonerDilation N M) (maximallyMixed (N+1)) = maximallyMixed (M+1) ∧
complementaryChannel (clonerDilation N M) (maximallyMixed (N+1)) = maximallyMixed (M-N+1) ∧
coherentInformation (receiverChannel (clonerDilation N M)) (maximallyMixed (N+1)) =
maximumCoherentInformation (clonerDilation N M) ∧
maximumCoherentInformation (clonerDilation N M) = Real.logb 2 (M+1) - Real.logb 2 ((M-N+1 : ℕ) : ℝ) ∧
maximumCoherentInformation (clonerDilation N M) = Real.logb 2 ((M+1 : ℝ) / (M-N+1 : ℕ)) ∧
∀ rho, IsDensityMatrix rho → coherentInformation (receiverChannel (clonerDilation N M)) rho ≤
coherentInformation (receiverChannel (clonerDilation N M)) (maximallyMixed (N+1)) := by sorry
end QAlgorithms.Papers.Independent.Support50.P01
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
For every pair of natural numbers , put and define by
for , , and , with natural-number subtraction truncated at zero. Let , , and , with applied by continuous functional calculus. Define
where , using the functional-calculus square root, and applies separately to each reference-indexed matrix block. Then is positive semidefinite with trace one, and
Furthermore, every positive semidefinite input matrix of trace one satisfies . The case is included: then and the displayed value is .
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.