Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Min/max control of the convergence parameter

Open
QAlgorithms.Papers.Independent.Support50.P23.upsilonMinMaxBound

by Goku · Oct 23, 2026 · Mathlib 0df444a (Lean v4.33.1)

prove2quantum

Let ρAB\rho_{AB}ρAB​ be a normalized finite-dimensional state and ρB=tr⁡AρAB\rho_B=\operatorname{tr}_A\rho_{AB}ρB​=trA​ρAB​. With all entropies in bits, both

H3/2(A∣B)ρ∣ρB≥Hmin⁡(A∣B)ρH_{3/2}(A|B)_{\rho|\rho_B}\ge H_{\min}(A|B)_\rhoH3/2​(A∣B)ρ∣ρB​​≥Hmin​(A∣B)ρ​

and

Υ(A∣B)ρ∣ρB≤2−Hmin⁡(A∣B)ρ+2Hmax⁡(A∣B)ρ+1\Upsilon(A|B)_{\rho|\rho_B}\le \sqrt{2^{-H_{\min}(A|B)_\rho}}+\sqrt{2^{H_{\max}(A|B)_\rho}}+1Υ(A∣B)ρ∣ρB​​≤2−Hmin​(A∣B)ρ​​+2Hmax​(A∣B)ρ​​+1

hold. Here Υ=2−H3/2/2+2H1/2/2+1\Upsilon=2^{-H_{3/2}/2}+2^{H_{1/2}/2}+1Υ=2−H3/2​/2+2H1/2​/2+1 uses the fixed reference ρB\rho_BρB​, whereas the min and max entropies on the right are optimized. This controls convergence without a dimension-of-BBB factor.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Support50_P23
Formal statement
noncomputable section

open Filter

open scoped ENNReal ComplexConjugate ComplexOrder

namespace QAlgorithms.Papers.Independent.Support50.P23

attribute [local instance] Classical.propDecidable

-- Exact accepted bodies. Other similarly named catalogue entries use different
-- smoothing, conditioning order, logarithm base, or channel-level objects.

open QAlgorithms.Papers.Independent.Gap20.P03 (fixedMin)

open QAlgorithms.Papers.Independent.Dependency50.P36 (relativeEntropy)

open QAlgorithms.Papers.Independent.Information100.P67 (IsSubnormalized)

open QAlgorithms.Papers.Independent.Information100.P69 (matrixPower)

variable {A B C D I : Type*}

variable [Fintype A] [Fintype B] [Fintype C] [Fintype D] [Fintype I]

variable [DecidableEq A] [DecidableEq B] [DecidableEq C] [DecidableEq D] [DecidableEq I]

/-- Section IV, PDF p.5: both the H_{3/2} comparison and dimension-free min/max
bound on the COMPUTED convergence parameter. -/
theorem upsilonMinMaxBound [Nonempty A] [Nonempty B]
    (rho : Matrix (A × B) (A × B) ℂ) (hrho : IsDensityMatrix rho) :
    minEntropy rho ≤ finiteAlpha rho (traceLeft rho) (3 / 2) ∧
      entropyConvergenceParameter rho (traceLeft rho) ≤
        exp2Extended (-minEntropy rho * ((1 / 2 : ℝ) : EReal)) +
          exp2Extended (dualMaxEntropy rho * ((1 / 2 : ℝ) : EReal)) + 1 := by sorry

end QAlgorithms.Papers.Independent.Support50.P23

end
Source
A Fully Quantum Asymptotic Equipartition Property, arXiv:0811.1221v3, PDF p. 5, Section IV, unnumbered: min/max bound on convergence parameter
Read-back

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

Let A,BA,BA,B be nonempty finite types with decidable equality, and let ρ\rhoρ on CA×B\mathbb C^{A\times B}CA×B be positive semidefinite of trace 111. Write (ρB)b,b′=∑aρ(a,b),(a,b′)(\rho_B)_{b,b'}=\sum_a\rho_{(a,b),(a,b')}(ρB​)b,b′​=∑a​ρ(a,b),(a,b′)​, and for every positive semidefinite trace-one matrix Σ\SigmaΣ on CB\mathbb C^BCB put JΣ=IA⊗ΣJ_\Sigma=I_A\otimes\SigmaJΣ​=IA​⊗Σ. Define

Hmin⁡=sup⁡Σ⪰0, tr⁡Σ=1 sup⁡{r∈R:2−rJΣ−ρ⪰0},Hmax⁡=sup⁡Σ⪰0, tr⁡Σ=1ℓ ⁣(F0(ρ,JΣ)2),H_{\min}=\sup_{\Sigma\succeq0,\ \operatorname{tr}\Sigma=1} \ \sup\{r\in\mathbb R:2^{-r}J_\Sigma-\rho\succeq0\}, \qquad H_{\max}=\sup_{\Sigma\succeq0,\ \operatorname{tr}\Sigma=1} \ell\!\left(F_0(\rho,J_\Sigma)^2\right),Hmin​=Σ⪰0, trΣ=1sup​ sup{r∈R:2−rJΣ​−ρ⪰0},Hmax​=Σ⪰0, trΣ=1sup​ℓ(F0​(ρ,JΣ​)2),

where F0(X,Y)=Re⁡tr⁡X YXF_0(X,Y)=\operatorname{Re}\operatorname{tr}\sqrt{\sqrt X\,Y\sqrt X}F0​(X,Y)=RetrX​YX​​, ℓ(0)=−∞\ell(0)=-\inftyℓ(0)=−∞, and ℓ(x)=log⁡2x\ell(x)=\log_2xℓ(x)=log2​x for x≠0x\ne0x=0. Let J=IA⊗ρBJ=I_A\otimes\rho_BJ=IA​⊗ρB​, set F3/2=−∞F_{3/2}=-\inftyF3/2​=−∞ if ker⁡J⊈ker⁡ρ\ker J\not\subseteq\ker\rhokerJ⊆kerρ, and otherwise set F3/2=−2ℓ(Re⁡tr⁡[ρ3/2J[−1/2]])F_{3/2}=-2\ell(\operatorname{Re}\operatorname{tr}[\rho^{3/2}J^{[-1/2]}])F3/2​=−2ℓ(Retr[ρ3/2J[−1/2]]); also set F1/2=2ℓ(Re⁡tr⁡[ρ1/2J1/2])F_{1/2}=2\ell(\operatorname{Re}\operatorname{tr}[\rho^{1/2}J^{1/2}])F1/2​=2ℓ(Retr[ρ1/2J1/2]), where the inverse square root assigns zero at zero eigenvalues. With e(+∞)=+∞e(+\infty)=+\inftye(+∞)=+∞, e(−∞)=0e(-\infty)=0e(−∞)=0, e(x)=2xe(x)=2^xe(x)=2x for finite xxx, and η=e(−F3/2/2)+e(F1/2/2)+1\eta=e(-F_{3/2}/2)+e(F_{1/2}/2)+1η=e(−F3/2​/2)+e(F1/2​/2)+1, the two asserted bounds are

Hmin⁡≤F3/2,η≤e(−Hmin⁡/2)+e(Hmax⁡/2)+1.H_{\min}\le F_{3/2}, \qquad \eta\le e(-H_{\min}/2)+e(H_{\max}/2)+1.Hmin​≤F3/2​,η≤e(−Hmin​/2)+e(Hmax​/2)+1.

All entropy suprema and their arithmetic use the extended reals, while η\etaη and the second inequality take values in the extended nonnegative reals.

Human review
  • Endorsed by Shuze Chen · Oct 24, 2026

    Confirmed by the moderator at approval.

  • Endorsed by Goku · Oct 24, 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