Theorem 3, p. 7 — for commuting ensembles the reversal operation's average entanglement fidelity is at least the square of any recovery's
ProvedQReversal.NearOpt.theorem_3Let be a trace-preserving completely positive map on a finite-dimensional system. Let be an ensemble of commuting density matrices — weights with , density matrices with — and let . Assume has full support, and let be the reversal operation. Then for every trace-preserving completely positive map ,
where is the average entanglement fidelity.
Consequently, if the best recovery has average entanglement fidelity , the reversal operation, which is computed from and alone, achieves at least : its error is at most twice the optimal error.
Formalization Note The recovery ranges over every finite family with , over every finite index type. Trace preservation of is the paper's noise model (§II); full support of is the standing assumption under which (12) is defined (p. 7). This is the statement of arXiv v2; the published J. Math. Phys. version proves it only for single states (Acknowledgments, p. 14).
import Mathlib import Definitions.Def_PeresTerno_kraus_basics import Definitions.Def_QReversal_NearOpt_Setting open Matrix PeresTerno open scoped MatrixOrder ComplexOrder
namespace QReversal.NearOpt
theorem theorem_3 {n L ι : Type*} [Fintype n] [DecidableEq n] [Fintype L] [Fintype ι]
(A : ι → Matrix n n ℂ) (hA : IsTracePreserving A)
(p : L → ℝ) (ρs : L → Matrix n n ℂ) (hE : IsCommutingEnsemble p ρs)
(ρ : Matrix n n ℂ) (hρ : ρ = avgState p ρs) (hfull : (krausUpdate A ρ).PosDef)
{κ : Type*} [Fintype κ] (R : κ → Matrix n n ℂ) (hR : IsTracePreserving R) :
Fbar p ρs (comp R A) ^ 2 ≤ Fbar p ρs (comp (reversal A ρ) A) := by sorry
end QReversal.NearOpt
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.