Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem I.18: liminf_(lambda→0+) Ra,Fbar(lambda)=liminf_(lambda→0+) Ra,Dbar(lambda)≥H(averageState).

Proved
QAlgorithms.Papers.Independent.Renew46.P43.weakSourceConverse

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

prove2quantum

For a finite pure-state iid source with average ρˉ\bar\rhoρˉ​, let Ra,Fˉ(λ)R_{a,\bar F}(\lambda)Ra,Fˉ​(λ) and Ra,Dˉ(λ)R_{a,\bar D}(\lambda)Ra,Dˉ​(λ) be the fixed-error limsups of the minimum arbitrary-encoding block rates under average overlap or half-trace-distance error, respectively. The error limit below is from positive values.

lim inf⁡λ↓0Ra,Fˉ(λ)=lim inf⁡λ↓0Ra,Dˉ(λ)≥H(ρˉ).\liminf_{\lambda\downarrow0}R_{a,\bar F}(\lambda)=\liminf_{\lambda\downarrow0}R_{a,\bar D}(\lambda)\ge H(\bar\rho).λ↓0liminf​Ra,Fˉ​(λ)=λ↓0liminf​Ra,Dˉ​(λ)≥H(ρˉ​).

This is the source result with its stated operational or information-theoretic domain.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Renew46_P43
Formal statement
noncomputable section

namespace QAlgorithms.Papers.Independent.Renew46.P43

open scoped BigOperators Matrix ComplexConjugate ComplexOrder ENNReal Matrix.Norms.L2Operator

open Filter MeasureTheory

open QAlgorithms.Papers.Independent.Information100.P08 (IsPure)

open QAlgorithms.Papers.Independent.Information100.P06 (Below)

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

open QAlgorithms.Papers.Independent.Dependency50.P19 (count both)

open QAlgorithms.Papers.Independent.Information100.P22 (coherentInformation)

attribute [local instance] Classical.propDecidable Classical.decEq

/-- Theorem I.18, PDF pp.18-19: one-sided error-to-zero liminf, not the value at zero. -/
theorem weakSourceConverse {a d : ℕ} (S : PureSource a d) :
    Filter.liminf (sourceRate S .arbitraryFidelity) (nhdsWithin 0 (Set.Ioi 0)) =
      Filter.liminf (sourceRate S .arbitraryDistortion) (nhdsWithin 0 (Set.Ioi 0)) ∧
    ENNReal.ofReal (vnEntropy S.average) ≤
      Filter.liminf (sourceRate S .arbitraryDistortion) (nhdsWithin 0 (Set.Ioi 0)) := by sorry

end QAlgorithms.Papers.Independent.Renew46.P43

end
Source
Coding Theorems of Quantum Information Theory, arXiv:quant-ph/9907077v1, PDF p. 18, Theorem I.18
Read-back

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

Let SSS consist of a>0a>0a>0 distinct pure density matrices ρx=∣vx⟩⟨vx∣\rho_x=|v_x\rangle\langle v_x|ρx​=∣vx​⟩⟨vx​∣ on Cd\mathbb C^dCd, with d>0d>0d>0, ∥vx∥=1\|v_x\|=1∥vx​∥=1, and a probability mass function pxp_xpx​ that may have zero weights, and write ρˉ=∑xpxρx\bar\rho=\sum_xp_x\rho_xρˉ​=∑x​px​ρx​. For every positive length ℓ\ellℓ, a code has a positive compressed dimension kkk, an arbitrary assignment of density matrices σw\sigma_wσw​ on Ck\mathbb C^kCk to all words www, and a decoder channel D\mathcal DD into (Cd)⊗ℓ(\mathbb C^d)^{\otimes\ell}(Cd)⊗ℓ; no channel extension of the encoder is required. A density matrix is positive semidefinite with trace one, and a channel has a finite Kraus representation Φ(X)=∑jKjXKj†\Phi(X)=\sum_jK_jXK_j^\daggerΦ(X)=∑j​Kj​XKj†​ with ∑jKj†Kj=I\sum_jK_j^\dagger K_j=I∑j​Kj†​Kj​=I. With pw=∏ipwip_w=\prod_ip_{w_i}pw​=∏i​pwi​​ and ρw=⨂iρwi\rho_w=\bigotimes_i\rho_{w_i}ρw​=⨂i​ρwi​​, put

Fc=∑wpw Re⁡Tr⁡(D(σw)ρw),Dc=∑wpw 12∥D(σw)−ρw∥1,∥X∥1=Re⁡Tr⁡X†X.F_c=\sum_wp_w\,\operatorname{Re}\operatorname{Tr}(\mathcal D(\sigma_w)\rho_w), \qquad D_c=\sum_wp_w\,\frac12\|\mathcal D(\sigma_w)-\rho_w\|_1, \quad \|X\|_1=\operatorname{Re}\operatorname{Tr}\sqrt{X^\dagger X}.Fc​=w∑​pw​ReTr(D(σw​)ρw​),Dc​=w∑​pw​21​∥D(σw​)−ρw​∥1​,∥X∥1​=ReTrX†X​.

For each real λ\lambdaλ, define RF(λ)R_F(\lambda)RF​(λ) by taking, at each length ℓ\ellℓ, the infimum in [0,∞][0,\infty][0,∞] of max⁡{log⁡2k/ℓ,0}\max\{\log_2k/\ell,0\}max{log2​k/ℓ,0} over codes with 1−Fc≤λ1-F_c\le\lambda1−Fc​≤λ, and then taking the limit superior along ℓ=n+1→∞\ell=n+1\to\inftyℓ=n+1→∞; define RD(λ)R_D(\lambda)RD​(λ) in the same way using Dc≤λD_c\le\lambdaDc​≤λ. Empty infima are +∞+\infty+∞. The limit inferiors of these two rate functions as λ\lambdaλ approaches zero through strictly positive real values satisfy

lim inf⁡λ→0+RF(λ)=lim inf⁡λ→0+RD(λ),max⁡{H(ρˉ),0}≤lim inf⁡λ→0+RD(λ),\liminf_{\lambda\to0^+}R_F(\lambda) = \liminf_{\lambda\to0^+}R_D(\lambda), \qquad \max\{H(\bar\rho),0\} \le \liminf_{\lambda\to0^+}R_D(\lambda),λ→0+liminf​RF​(λ)=λ→0+liminf​RD​(λ),max{H(ρˉ​),0}≤λ→0+liminf​RD​(λ),

where H(ρ)=−Re⁡Tr⁡f(ρ)H(\rho)=-\operatorname{Re}\operatorname{Tr}f(\rho)H(ρ)=−ReTrf(ρ) with f(t)=tlog⁡2tf(t)=t\log_2tf(t)=tlog2​t defined by continuous functional calculus, and both limit inferiors and the inequalities are interpreted in [0,∞][0,\infty][0,∞].

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