Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Product-input smooth-entropy relaxation

Open
QAlgorithms.Papers.Independent.Support50.P04.product_smooth_relaxation

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

prove2quantum

Let N:A→B\mathcal N:A\to BN:A→B be a CPTP channel on finite positive-dimensional complex Hilbert spaces with fixed orthonormal bases. Write R^(n;ε)\widehat R(n;\varepsilon)R(n;ε) and R^cpp(n;ε)\widehat R^{\mathrm{cpp}}(n;\varepsilon)Rcpp(n;ε) for the maximum unassisted and classical-post-processing assisted entanglement-transmission rates. A code has a positive integer message dimension MMM, rate log⁡2M/n\log_2 M/nlog2​M/n, and squared fidelity to the normalized maximally entangled target at least 1−ε1-\varepsilon1−ε. A cpp encoder may retain sender memory; its decoder is LOCC across that memory and the receiver outputs, with its final quantum output at the receiver and no preshared entanglement. Use one finite-dimensional isometry U:A↪B⊗EU:A\hookrightarrow B\otimes EU:A↪B⊗E and its receiver/complementary channels. Let n≥1n\ge1n≥1, 0<ε<10<\varepsilon<10<ε<1, and 1/n<ε1/\sqrt n<\varepsilon1/n​<ε. For a normalized center ω\omegaω, define Hmin⁡δ(A∣E)ωH_{\min}^{\delta}(A|E)_\omegaHminδ​(A∣E)ω​ as the supremum of real λ\lambdaλ for which there exist ωˉ⪰0\bar\omega\succeq0ωˉ⪰0 of trace at most one and a normalized density σE\sigma_EσE​ such that F(ωˉ,ω)≥1−δ2F(\bar\omega,\omega)\ge1-\delta^2F(ωˉ,ω)≥1−δ2 and ωˉ⪯2−λIA⊗σE\bar\omega\preceq2^{-\lambda}I_A\otimes\sigma_Eωˉ⪯2−λIA​⊗σE​. Supremums are taken in the extended reals. For each normalized single-use density input ρ\rhoρ, put ωρ=(id⊗Nc)(ψρ)\omega^\rho=(\mathrm{id}\otimes\mathcal N^c)(\psi^\rho)ωρ=(id⊗Nc)(ψρ) and δn=ε−1/n\delta_n=\sqrt{\varepsilon-1/\sqrt n}δn​=ε−1/n​​. Then

R^(n;ε)≥sup⁡ρ∈S(A)Hmin⁡δn(An∣En)(ωρ)⊗n−2log⁡2n−1n.\widehat R(n;\varepsilon)\ge\sup_{\rho\in\mathcal S(A)}\frac{H_{\min}^{\delta_n}(A^n|E^n)_{(\omega^\rho)^{\otimes n}}-2\log_2 n-1}{n}.R(n;ε)≥ρ∈S(A)sup​nHminδn​​(An∣En)(ωρ)⊗n​−2log2​n−1​.

This is the product-input relaxation of the optimized inner bound.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Support50_P04
Formal statement
noncomputable section

open scoped Matrix ComplexConjugate ComplexOrder

open QAlgorithms

namespace QAlgorithms.Papers.Independent.Support50.P04

variable {A B E : Type} [Fintype A] [DecidableEq A]

variable [Fintype B] [DecidableEq B] [Fintype E] [DecidableEq E]

/-- arXiv:1504.04617v2, PDF p.14, Eq.(47).
Printed delta=sqrt(epsilon)-1/sqrt(n), n>1/epsilon; changed by the paragraph's
eta=1/sqrt(n) substitution in (46) to delta=sqrt(epsilon-1/sqrt(n)), 1/sqrt(n)<epsilon. -/
theorem product_smooth_relaxation [Nonempty A] [Nonempty B] [Nonempty E]
    (U : Matrix (B × E) A ℂ) (hU : Uᴴ * U = 1)
    (n : ℕ) (hn : 0 < n) (epsilon : ℝ) (he0 : 0 < epsilon) (he1 : epsilon < 1)
    (hlarge : 1/Real.sqrt (n : ℝ) < epsilon) :
    productSmoothInner (QAlgorithms.Papers.Independent.Information100.P67.complementaryChannel U) n epsilon ≤
      unassistedBoundary (QAlgorithms.Papers.Independent.Information100.P67.receiverChannel U) n epsilon := by sorry

end QAlgorithms.Papers.Independent.Support50.P04

end
Source
Quantum Coding with Finite Resources, arXiv:1504.04617v2, PDF p. 14, Eq. (47)
Read-back

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

Let A,B,EA,B,EA,B,E be nonempty finite sets with decidable equality, let U:CA→CB×EU:\mathbb C^A\to\mathbb C^{B\times E}U:CA→CB×E satisfy U†U=IAU^\dagger U=I_AU†U=IA​, and put N(X)=Tr⁡E(UXU†)\mathcal N(X)=\operatorname{Tr}_E(UXU^\dagger)N(X)=TrE​(UXU†) and Nc(X)=Tr⁡B(UXU†)\mathcal N^c(X)=\operatorname{Tr}_B(UXU^\dagger)Nc(X)=TrB​(UXU†). For every integer n>0n>0n>0 and every 0<ϵ<10<\epsilon<10<ϵ<1 satisfying 1/n<ϵ1/\sqrt n<\epsilon1/n​<ϵ, the assertion is

sup⁡ρ≥0, Tr⁡ρ=1Hmin⁡ϵ−1/n ⁣(ΩNc,ρ⊗n)−2log⁡2n−1n≤Cu(N,n,ϵ).\sup_{\rho\ge0,\ \operatorname{Tr}\rho=1} \frac{ H_{\min}^{\sqrt{\epsilon-1/\sqrt n}} \!\left(\Omega_{\mathcal N^c,\rho}^{\otimes n}\right) -2\log_2n-1}{n} \le C_{\mathrm u}(\mathcal N,n,\epsilon).ρ≥0, Trρ=1sup​nHminϵ−1/n​​​(ΩNc,ρ⊗n​)−2log2​n−1​≤Cu​(N,n,ϵ).

The input ρ\rhoρ on the left is a single-copy matrix on AAA; writing S=ρS=\sqrt\rhoS=ρ​, define (Pρ)(i,j),(k,l)=SijSkl‾(P_\rho)_{(i,j),(k,l)}=S_{ij}\overline{S_{kl}}(Pρ​)(i,j),(k,l)​=Sij​Skl​​ and ΩNc,ρ=(id⊗Nc)(Pρ)\Omega_{\mathcal N^c,\rho}=(\mathrm{id}\otimes\mathcal N^c)(P_\rho)ΩNc,ρ​=(id⊗Nc)(Pρ​). The displayed tensor power groups the nnn reference factors together and the nnn environment factors together, with entries equal to products of single-copy entries. For a matrix ω\omegaω on I×JI\times JI×J,

Hmin⁡δ(ω)=sup⁡{λ∈R:∃ τ≥0, Re⁡Tr⁡τ≤1, F(τ,ω)≥1−δ2, ∃ σ≥0, Tr⁡σ=1, τ≤2−λII⊗σ},H_{\min}^{\delta}(\omega)= \sup\{\lambda\in\mathbb R: \exists\,\tau\ge0,\ \operatorname{Re}\operatorname{Tr}\tau\le1,\ F(\tau,\omega)\ge1-\delta^2,\ \exists\,\sigma\ge0,\ \operatorname{Tr}\sigma=1,\ \tau\le2^{-\lambda}I_I\otimes\sigma\},Hminδ​(ω)=sup{λ∈R:∃τ≥0, ReTrτ≤1, F(τ,ω)≥1−δ2, ∃σ≥0, Trσ=1, τ≤2−λII​⊗σ},

as an extended-real supremum, where σ\sigmaσ acts on JJJ, F(X,Y)=(Re⁡Tr⁡X Y X)2F(X,Y)=(\operatorname{Re}\operatorname{Tr}\sqrt{\sqrt X\,Y\,\sqrt X})^2F(X,Y)=(ReTrX​YX​​)2, and zero τ\tauτ is allowed. The quantity CuC_{\mathrm u}Cu​ is the extended-real supremum of rates RRR for which positive MMM and channel maps encoding MMM dimensions into AnA^nAn and decoding BnB^nBn into MMM dimensions give log⁡2M/n≥R\log_2M/n\ge Rlog2​M/n≥R and entanglement fidelity at least 1−ϵ1-\epsilon1−ϵ after N⊗n\mathcal N^{\otimes n}N⊗n. Entanglement fidelity is the overlap of the output on half of the normalized maximally entangled state of dimension MMM with that same state. Channels are finite complete Kraus sums. The additional hypothesis excludes every nnn with n≤ϵ−2n\le\epsilon^{-2}n≤ϵ−2, in particular n=1n=1n=1.

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