Section 4.3 / (51)–(53): high mean fidelity without norm convergence
ProvedQAlgorithms.Papers.Independent.Gap20.P07.meanFidelity_counterexampleFor each integer , let , , and define the CPTP pinching channel . Then
Thus no function with as can give uniformly over all positive dimensions and CPTP endomorphisms. This counterexample does not contradict equality of capacities, where logical subspaces may be selected.
import Definitions.Def_QAlgorithms_Papers_Independent_Gap20_P07
noncomputable section
open Filter MeasureTheory
open scoped Topology ENNReal NNReal Matrix ComplexOrder ComplexConjugate
attribute [local instance] Classical.propDecidable
attribute [local instance] Classical.decEq
attribute [local instance] IsStarNormal.instContinuousFunctionalCalculus
namespace QAlgorithms.Papers.Independent.Gap20.P07
open QAlgorithms
open QAlgorithms.RealRegisters (Hilbert)
open QAlgorithms.Papers.Independent.Dependency50.P33
(State IsTraceClass IsTraceClassChannel traceSize rankOne tensorApply)
open QAlgorithms.Papers.Independent.Frontier30.P14
(Operation isometryCoeff conjugateKernel)
open QAlgorithms.Papers.Independent.Claimed50.P35
(stabilizedTraceClassNorm traceClassTensorMap traceClassChannelPower
transpose transposeBound TraceLinear tracePair MeasurePrepare)
open QAlgorithms.Papers.Independent.Gap48.P29 (fe FiniteCPMap)
/-- PDF p.16, (51)--(53): the actual firstCoordinatePinching family has high mean/channel
fidelity and ordinary-norm distance bounded below by one. The two-dimensional
superposition gives off-diagonal eigenvalues +1/2 and -1/2, hence full trace norm one. -/
theorem meanFidelity_counterexample :
(∀ d : ℕ+, 2 ≤ (d : ℕ) → IsChannel (firstCoordinatePinching d) ∧
QuditFidelity.entanglementFidelity d d.pos (firstCoordinatePinching d) =
((d : ℝ)^2 - 2 * d + 2) / (d : ℝ)^2 ∧
1 ≤ ordinaryNorm d (fun X => firstCoordinatePinching d X - X)) ∧
Tendsto (fun n : ℕ => QuditFidelity.averageFidelity (n + 2) (by omega)
(firstCoordinatePinching ⟨n + 2, by omega⟩)) atTop (𝓝 1) ∧
Tendsto (fun n : ℕ => QuditFidelity.entanglementFidelity (n + 2) (by omega)
(firstCoordinatePinching ⟨n + 2, by omega⟩)) atTop (𝓝 1) ∧
¬ ∃ f : ℝ → ℝ, Tendsto f (𝓝[<] (1 : ℝ)) (𝓝 0) ∧
∀ (d : ℕ) (hd : 0 < d) (T : Superop (Fin d) (Fin d)), IsChannel T →
ordinaryNorm d (fun X => T X - X) ≤ f (QuditFidelity.entanglementFidelity d hd T) := by sorry
end QAlgorithms.Papers.Independent.Gap20.P07
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
For each positive integer , let and define . For a map write , with ; define as the integral of against the pushforward of normalized Haar measure under , and define , where , , and the amplification acts on the second factor blockwise. The declaration asserts jointly that, for every , admits a finite Kraus representation whose Kraus operators satisfy ,
that both and tend to as through the natural numbers; and that there is no function with such that, for every positive integer and every finite Kraus channel on ,
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.