Quantum Stein lemma at every fixed positive error
OpenQAlgorithms.Papers.Independent.Gap48.P41.quantum_SteinLet be a nonempty finite orthonormal basis, and let be positive semidefinite complex matrices of trace one with . For a positive integer , an arbitrary effect accepts the null state . Write and . All logarithms are natural, with in rates. Write , using support-restricted spectral logarithms inside the trace.
For each fixed , let be the minimum of over all effects with . This minimum is attained, and
This identifies the optimal type-II exponential rate at a fixed nonzero type-I tolerance.
Formalization Note The printed is corrected to : the zero-error endpoint is false for full-rank unequal states. The ordinary limit uses .
import Definitions.Def_QAlgorithms_Papers_Independent_Gap48_P41
noncomputable section
namespace QAlgorithms.Papers.Independent.Gap48.P41
open scoped Matrix ComplexOrder BigOperators
open Filter QAlgorithms
open QAlgorithms.Papers.Independent.Information100.P60 (Mat copies)
open QAlgorithms.Papers.Independent.Information100.P71 (supportPower)
open QAlgorithms.Papers.Independent.Information100.P07 (positiveProjection)
open QAlgorithms.Papers.Independent.Dependency50.P28
(tr SupportLE relativeEntropy scalarPower classicalD typeI accept TestSequence
scaled logNonnegativeExtended)
open QAlgorithms.Papers.Independent.Claimed50.P15 (legendre)
open QAlgorithms.Papers.Independent.Information100.P56 (IsLaw)
attribute [local instance] Classical.propDecidable Classical.decEq
variable {I : Type} [Fintype I] [DecidableEq I]
/-- PDF pp.4--5, Theorem 2, (10): ordinary fixed-positive-error quantum Stein limit,
including minimum attainment. SOURCE CORRECTION: printed 0<=eps<1 is changed
to 0<eps<1. For rho=diag(1/2,1/2), sigma=diag(3/4,1/4), eps=0 forces A=I,
so the rate is 0 while D=log(4/3)/2>0. No zero-error direct assertion is made. -/
theorem quantum_Stein [Nonempty I]
(rho sigma : Mat I) (hrho : IsDensityMatrix rho) (hsigma : IsDensityMatrix sigma)
(hsupp : SupportLE rho sigma) (eps : ℝ) (heps0 : 0 < eps) (heps1 : eps < 1) :
(∀ k : ℕ, IsLeast (betaValues rho sigma (k + 1) eps)
(betaStar rho sigma (k + 1) eps)) ∧
Tendsto (fun k => scaled k (logNonnegativeExtended (betaStar rho sigma (k + 1) eps)))
atTop (nhds (-relativeEntropy rho sigma)) := by sorry
end QAlgorithms.Papers.Independent.Gap48.P41
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
Let be a nonempty finite type with decidable equality, let be positive semidefinite complex matrices indexed by , each of trace , and assume . Let . For each natural number , define
where is indexed by and . Put for and otherwise, and let
where matrix logarithms use spectral functional calculus with . Then, for every , belongs to and is no greater than any member of that set, and
in the extended real line. Thus the assertion concerns every positive block length, beginning with .
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.