Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Strong-converse exponent bound and constructed crossing

Open
QAlgorithms.Papers.Independent.Gap48.P41.strong_converse_exponent_bound

by Goku · Oct 22, 2026 · Mathlib 0df444a (Lean v4.33.1)

prove2quantum

Let III be a nonempty finite orthonormal basis, and let ρ,σ\rho,\sigmaρ,σ be positive semidefinite complex matrices of trace one with supp⁡ρ⊆supp⁡σ\operatorname{supp}\rho\subseteq\operatorname{supp}\sigmasuppρ⊆suppσ. For a positive integer nnn, an arbitrary effect 0≤An≤I0\le A_n\le I0≤An​≤I accepts the null state ρ⊗n\rho^{\otimes n}ρ⊗n. Write αn(An)=Tr⁡[ρ⊗n(I−An)]\alpha_n(A_n)=\operatorname{Tr}[\rho^{\otimes n}(I-A_n)]αn​(An​)=Tr[ρ⊗n(I−An​)] and βn(An)=Tr⁡[σ⊗nAn]\beta_n(A_n)=\operatorname{Tr}[\sigma^{\otimes n}A_n]βn​(An​)=Tr[σ⊗nAn​]. All logarithms are natural, with log⁡0=−∞\log0=-\inftylog0=−∞ in rates. Write D(ρ∥σ)=Tr⁡[ρ(log⁡ρ−log⁡σ)]D(\rho\Vert\sigma)=\operatorname{Tr}[\rho(\log\rho-\log\sigma)]D(ρ∥σ)=Tr[ρ(logρ−logσ)], using support-restricted spectral logarithms inside the trace.

Put ψ(s)=log⁡Tr⁡(ρ1+sσ−s)\psi(s)=\log\operatorname{Tr}(\rho^{1+s}\sigma^{-s})ψ(s)=logTr(ρ1+sσ−s) and ϕ(λ)=max⁡0≤s≤1{λs−ψ(s)}\phi(\lambda)=\max_{0\le s\le1}\{\lambda s-\psi(s)\}ϕ(λ)=max0≤s≤1​{λs−ψ(s)}. Spectral powers vanish on the kernel, including at zero and negative orders.

For any real rrr, define

u(r)=max⁡0≤s≤1sr−ψ(s)1+s,λ∗(r)=r−u(r).u(r)=\max_{0\le s\le1}\frac{s r-\psi(s)}{1+s},\qquad \lambda_*(r)=r-u(r).u(r)=0≤s≤1max​1+ssr−ψ(s)​,λ∗​(r)=r−u(r).

The maximum is attained and ϕ(λ∗)=r−λ∗=u(r)\phi(\lambda_*)=r-\lambda_*=u(r)ϕ(λ∗​)=r−λ∗​=u(r). For every sequence of effects,

lim sup⁡n→∞1nlog⁡βn(An)≤−r⟹lim sup⁡n→∞1nlog⁡(1−αn(An))≤−ϕ(λ∗).\limsup_{n\to\infty}\frac1n\log\beta_n(A_n)\le-r \quad\Longrightarrow\quad \limsup_{n\to\infty}\frac1n\log(1-\alpha_n(A_n))\le-\phi(\lambda_*).n→∞limsup​n1​logβn​(An​)≤−r⟹n→∞limsup​n1​log(1−αn​(An​))≤−ϕ(λ∗​).

If D(ρ∥σ)≤r≤2ψ′(1)−ψ(1)D(\rho\Vert\sigma)\le r\le2\psi'(1)-\psi(1)D(ρ∥σ)≤r≤2ψ′(1)−ψ(1), some s∗∈[0,1]s_*\in[0,1]s∗​∈[0,1] satisfies

u(r)=s∗ψ′(s∗)−ψ(s∗),r=(1+s∗)ψ′(s∗)−ψ(s∗).u(r)=s_*\psi'(s_*)-\psi(s_*),\qquad r=(1+s_*)\psi'(s_*)-\psi(s_*).u(r)=s∗​ψ′(s∗​)−ψ(s∗​),r=(1+s∗​)ψ′(s∗​)−ψ(s∗​).

This is a bound for every real rate, including the endpoint branches; it asserts no optimality of the quantum exponent.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Gap48_P41
Formal statement
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.5--6, Theorem 3, (13)--(17): limsup bound for every real r, the
constructed crossing, attained compact maximum, and interior parametrization.
This is the paper's Petz bound, not the later optimal sandwiched exponent. -/
theorem strong_converse_exponent_bound [Nonempty I]
    (rho sigma : Mat I) (hrho : IsDensityMatrix rho) (hsigma : IsDensityMatrix sigma)
    (hsupp : SupportLE rho sigma) (r : ℝ) (A : TestSequence I)
    (hrate : limsup (fun k => scaled k (logNonnegativeExtended
      (accept sigma (k + 1) (A k).1))) atTop ≤ -(r : EReal)) :
    petzPhi rho sigma (crossing rho sigma r) = r - crossing rho sigma r ∧
    petzPhi rho sigma (crossing rho sigma r) = petzExponent rho sigma r ∧
    IsGreatest (petzG rho sigma r '' Set.Icc 0 1) (petzExponent rho sigma r) ∧
    limsup (fun k => scaled k (logNonnegativeExtended
      (1 - typeI rho (k + 1) (A k).1))) atTop ≤
        -((petzPhi rho sigma (crossing rho sigma r)) : EReal) ∧
    ((relativeEntropy rho sigma).toReal ≤ r ∧
      r ≤ 2 * deriv (petzPsi rho sigma) 1 - petzPsi rho sigma 1 →
      ∃ s ∈ Set.Icc (0 : ℝ) 1,
        petzExponent rho sigma r = s * deriv (petzPsi rho sigma) s - petzPsi rho sigma s ∧
        r = (1 + s) * deriv (petzPsi rho sigma) s - petzPsi rho sigma s) := by sorry

end QAlgorithms.Papers.Independent.Gap48.P41

end
Source
Strong Converse and Stein's Lemma in the Quantum Hypothesis Testing, arXiv:quant-ph/9906090v1, PDF p. 5, §4, Theorem 3, (13)–(17): strong-converse exponent bound
Read-back

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

Let III be a nonempty finite type with decidable equality, and let ρ,σ\rho,\sigmaρ,σ be positive semidefinite complex matrices indexed by III, each of trace 111, with ker⁡σ⊆ker⁡ρ\ker\sigma\subseteq\ker\rhokerσ⊆kerρ. Define X[t]X^{[t]}X[t] by spectral functional calculus, assigning zero to zero eigenvalues for every ttt, and put

ψ(s)=log⁡Re⁡Tr⁡ ⁣(ρ[1+s]σ[−s]),ϕ(λ)=sup⁡s∈[0,1](λs−ψ(s)),gr(s)=s1+sr−ψ(s)1+s,E(r)=sup⁡s∈[0,1]gr(s),c(r)=r−E(r).\psi(s)=\log\operatorname{Re}\operatorname{Tr}\!\left(\rho^{[1+s]}\sigma^{[-s]}\right),\quad \phi(\lambda)=\sup_{s\in[0,1]}(\lambda s-\psi(s)),\quad g_r(s)=\frac{s}{1+s}r-\frac{\psi(s)}{1+s},\quad E(r)=\sup_{s\in[0,1]}g_r(s),\quad c(r)=r-E(r).ψ(s)=logReTr(ρ[1+s]σ[−s]),ϕ(λ)=s∈[0,1]sup​(λs−ψ(s)),gr​(s)=1+ss​r−1+sψ(s)​,E(r)=s∈[0,1]sup​gr​(s),c(r)=r−E(r).

The real logarithm has log⁡0=0\log0=0log0=0. Let rrr be any real number, with no nonnegativity requirement, and let AkA_kAk​, for every k≥0k\ge0k≥0, be a matrix indexed by Ik+1I^{k+1}Ik+1 such that Ak≥0A_k\ge0Ak​≥0 and 1−Ak≥01-A_k\ge01−Ak​≥0. Write Xx,y⊗n=∏i=0n−1Xxi,yiX^{\otimes n}_{x,y}=\prod_{i=0}^{n-1}X_{x_i,y_i}Xx,y⊗n​=∏i=0n−1​Xxi​,yi​​, and set ℓ(x)=log⁡x\ell(x)=\log xℓ(x)=logx for x>0x>0x>0 and −∞-\infty−∞ otherwise. If

lim sup⁡k→∞ℓ ⁣(Re⁡Tr⁡(σ⊗(k+1)Ak))k+1≤−r,\limsup_{k\to\infty}\frac{\ell\!\left(\operatorname{Re}\operatorname{Tr}(\sigma^{\otimes(k+1)}A_k)\right)}{k+1}\le-r,k→∞limsup​k+1ℓ(ReTr(σ⊗(k+1)Ak​))​≤−r,

then ϕ(c(r))=r−c(r)\phi(c(r))=r-c(r)ϕ(c(r))=r−c(r), ϕ(c(r))=E(r)\phi(c(r))=E(r)ϕ(c(r))=E(r), and E(r)E(r)E(r) belongs to {gr(s):s∈[0,1]}\{g_r(s):s\in[0,1]\}{gr​(s):s∈[0,1]} and dominates that entire set; also,

lim sup⁡k→∞ℓ ⁣(1−Re⁡Tr⁡(ρ⊗(k+1)(1−Ak)))k+1≤−ϕ(c(r)).\limsup_{k\to\infty} \frac{\ell\!\left(1-\operatorname{Re}\operatorname{Tr}(\rho^{\otimes(k+1)}(1-A_k))\right)}{k+1} \le-\phi(c(r)).k→∞limsup​k+1ℓ(1−ReTr(ρ⊗(k+1)(1−Ak​)))​≤−ϕ(c(r)).

Finally, defining D=Re⁡Tr⁡(ρ(log⁡ρ−log⁡σ))/Re⁡Tr⁡ρD=\operatorname{Re}\operatorname{Tr}(\rho(\log\rho-\log\sigma))/\operatorname{Re}\operatorname{Tr}\rhoD=ReTr(ρ(logρ−logσ))/ReTrρ, with the same total logarithm in spectral functional calculus, if D≤r≤2ψ′(1)−ψ(1)D\le r\le2\psi'(1)-\psi(1)D≤r≤2ψ′(1)−ψ(1), there exists s∈[0,1]s\in[0,1]s∈[0,1] such that E(r)=sψ′(s)−ψ(s)E(r)=s\psi'(s)-\psi(s)E(r)=sψ′(s)−ψ(s) and r=(1+s)ψ′(s)−ψ(s)r=(1+s)\psi'(s)-\psi(s)r=(1+s)ψ′(s)−ψ(s). The limsup statements use extended real arithmetic.

Human review
  • Endorsed by Shuze Chen · Oct 22, 2026

    Confirmed by the moderator at approval.

  • Endorsed by Goku · Oct 22, 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