Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

§5.6, quantum Sanov theorem

Open
QAlgorithms.Papers.Independent.Intakenext100.P45.quantumSanov

by Gohan · Oct 31, 2026 · Mathlib 0df444a (Lean v4.33.1)

prove2quantum

Let Γ\GammaΓ be any family of densities on a fixed finite-dimensional space, and let σ\sigmaσ be a density. There exists one sequence of orthogonal projections such that αn(ρ)→0\alpha_n(\rho)\to0αn​(ρ)→0 for every ρ∈Γ\rho\in\Gammaρ∈Γ, and

βR=inf⁡ρ∈ΓS(ρ∥σ).\beta_R=\inf_{\rho\in\Gamma}S(\rho\|\sigma).βR​=ρ∈Γinf​S(ρ∥σ).

Every sequence of binary effects with the same pointwise vanishing type-I errors has upper type-II rate at most this infimum. No uniform convergence or compactness assumption is imposed. For Γ=∅\Gamma=\varnothingΓ=∅, the extended infimum is +∞+\infty+∞ and the null constraint is vacuous.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Intakenext100_P45
Formal statement
noncomputable section

namespace QAlgorithms.Papers.Independent.Intakenext100.P45

open QAlgorithms Filter

open scoped Matrix ComplexConjugate ComplexOrder ENNReal BigOperators Topology

open QAlgorithms.Papers.Independent.Information100.P07
  (support positivePart positiveProjection IsProjection)

open QAlgorithms.Papers.Independent.Information100.P56 (IsLaw)

open QAlgorithms.Papers.Independent.Information100.P60 (Mat copies)

open QAlgorithms.Papers.Independent.Dependency50.P28 (logNonnegativeExtended relativeEntropy)

open QAlgorithms.Papers.Independent.Claimed50.P15 (positiveMatrixRpow)

attribute [local instance] Classical.propDecidable Classical.decEq

variable {I J : Type} [Fintype I] [DecidableEq I] [Fintype J] [DecidableEq J]

/-- arXiv:0708.4282v1, PDF p.31: §5.6, quantum Sanov theorem. -/
theorem quantumSanov (Gamma : Set (Mat I)) (hGamma : ∀ rho ∈ Gamma, IsDensityMatrix rho)
    (sigma : Mat I) (hsigma : IsDensityMatrix sigma) :
    let D : EReal := sInf {v | ∃ rho ∈ Gamma, v=relativeEntropy rho sigma}
    (∃ P : QuantumTests I, (∀ k, IsProjection (P k)) ∧
      (∀ rho ∈ Gamma, Tendsto (quantumAlpha rho P) atTop (nhds 0)) ∧
      HasRate (quantumBeta sigma P) D) ∧
    (∀ P : QuantumTests I, (∀ k, IsEffect (P k)) →
      (∀ rho ∈ Gamma, Tendsto (quantumAlpha rho P) atTop (nhds 0)) →
      Filter.limsup (fun k => rateTerm (quantumBeta sigma P k) k) atTop ≤ D) := by sorry

end QAlgorithms.Papers.Independent.Intakenext100.P45

end
Source
Asymptotic Error Rates in Quantum Hypothesis Testing, arXiv:0708.4282v1, PDF p. 31, §5.6, quantum Sanov theorem
Read-back

What the Lean code literally says, in plain math · gpt-6.1-sol

Let III be any finite index type with decidable equality, let Γ\GammaΓ be any set of complex matrices indexed by III, assume every ρ∈Γ\rho\in\Gammaρ∈Γ is positive semidefinite with trace 111, and let σ\sigmaσ also be positive semidefinite with trace 111. Put D=inf⁡ρ∈ΓD(ρ∥σ)D=\inf_{\rho\in\Gamma}D(\rho\Vert\sigma)D=infρ∈Γ​D(ρ∥σ), where D(ρ∥σ)D(\rho\Vert\sigma)D(ρ∥σ) is +∞+\infty+∞ if ker⁡σ⊈ker⁡ρ\ker\sigma\not\subseteq\ker\rhokerσ⊆kerρ, and otherwise is Re⁡Tr⁡(ρ(log⁡ρ−log⁡σ))/Re⁡Tr⁡ρ\operatorname{Re}\operatorname{Tr}(\rho(\log\rho-\log\sigma))/\operatorname{Re}\operatorname{Tr}\rhoReTr(ρ(logρ−logσ))/ReTrρ, with natural logarithms, matrix logarithms defined by continuous functional calculus, and scalar log⁡0=0\log0=0log0=0. For any sequence P=(Pk)k≥0P=(P_k)_{k\ge0}P=(Pk​)k≥0​ of matrices on Ik+1I^{k+1}Ik+1, define

αk(ρ,P)=Re⁡Tr⁡(Pkρ⊗(k+1)),βk(σ,P)=Re⁡Tr⁡((1−Pk)σ⊗(k+1)),\alpha_k(\rho,P)=\operatorname{Re}\operatorname{Tr}(P_k\rho^{\otimes(k+1)}), \qquad \beta_k(\sigma,P)=\operatorname{Re}\operatorname{Tr}((1-P_k)\sigma^{\otimes(k+1)}),αk​(ρ,P)=ReTr(Pk​ρ⊗(k+1)),βk​(σ,P)=ReTr((1−Pk​)σ⊗(k+1)),

where tensor powers have entries given by products of the corresponding matrix entries. There exists such a sequence in which every PkP_kPk​ is Hermitian and satisfies Pk2=PkP_k^2=P_kPk2​=Pk​, for which αk(ρ,P)→0\alpha_k(\rho,P)\to0αk​(ρ,P)→0 for every individual ρ∈Γ\rho\in\Gammaρ∈Γ, and

−L(βk(σ,P))k+1⟶D.\frac{-L(\beta_k(\sigma,P))}{k+1}\longrightarrow D.k+1−L(βk​(σ,P))​⟶D.

Moreover, every sequence with Pk≥0P_k\ge0Pk​≥0 and 1−Pk≥01-P_k\ge01−Pk​≥0 for all kkk, and with αk(ρ,P)→0\alpha_k(\rho,P)\to0αk​(ρ,P)→0 for every ρ∈Γ\rho\in\Gammaρ∈Γ, satisfies

lim sup⁡k→∞−L(βk(σ,P))k+1≤D.\limsup_{k\to\infty}\frac{-L(\beta_k(\sigma,P))}{k+1}\le D.k→∞limsup​k+1−L(βk​(σ,P))​≤D.

Here L(x)=log⁡xL(x)=\log xL(x)=logx for x>0x>0x>0 and L(x)=−∞L(x)=-\inftyL(x)=−∞ otherwise, and the limits and infimum take values in the extended real line. The convergence over Γ\GammaΓ is pointwise, with no uniformity condition. The set Γ\GammaΓ may be empty; then D=+∞D=+\inftyD=+∞ and its universally quantified conditions are vacuous. The hypothesis on σ\sigmaσ excludes empty III.

Human review
  • Endorsed by Shuze Chen · Oct 31, 2026

    Confirmed by the moderator at approval.

  • Endorsed by Gohan · Oct 31, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me