Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

§6.4.2 (6.97)–(6.100): product-state finite-size bound

Open
QAlgorithms.Papers.Independent.Intakenext100.P07.product_finite_size_bound

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

prove2quantum

All quantum systems below are finite-dimensional complex Hilbert spaces of positive dimension, and logarithms are binary. A normalized state is positive semidefinite with trace one; a subnormalized state has trace at most one. Marginals are partial traces. Here FFF is squared algebraic fidelity, F∗F_*F∗​ is squared generalized fidelity, P=1−F∗P=\sqrt{1-F_*}P=1−F∗​​ is purified distance, and Δ(ρ,τ)=max⁡0≤M≤I∣Tr⁡((ρ−τ)M)∣\Delta(\rho,\tau)=\max_{0\le M\le I}|\operatorname{Tr}((\rho-\tau)M)|Δ(ρ,τ)=max0≤M≤I​∣Tr((ρ−τ)M)∣. The symbols Hmin⁡H_{\min}Hmin​ and Hmax⁡H_{\max}Hmax​ denote conditional min- and max-entropy, with values +∞+\infty+∞ and −∞-\infty−∞ at zero. Smoothing optimizes over all subnormalized states in the purified-distance ball; every entropy radius satisfies 0≤ε<Tr⁡ρ0\le\varepsilon<\sqrt{\operatorname{Tr}\rho}0≤ε<Trρ​. Write g(t)=−log⁡2(1−1−t2)g(t)=-\log_2(1-\sqrt{1-t^2})g(t)=−log2​(1−1−t2​) for 0<t<10<t<10<t<1.

For a finite set of normalized supported pairs (ρ_i,σ_i), choose a common C≥0 before all n and all sequences of pairs. For n≥1, (1/n)Dmax^ε(⊗ρ_i||⊗σ_i)≤(1/n)ΣD_i + (g(ε)+(1/(2log(e)n))ΣV_i)/sqrt(n)+C/n.

1nDmax⁡ε(⨂iρi∥⨂iσi)≤1n∑iDi+1n(g(ε)+∑iVi2nlog⁡2e)+Cn.\frac1nD_{\max}^\varepsilon(\bigotimes_i\rho_i\|\bigotimes_i\sigma_i)\le\frac1n\sum_iD_i+\frac1{\sqrt n}\left(g(\varepsilon)+\frac{\sum_iV_i}{2n\log_2e}\right)+\frac Cn.n1​Dmaxε​(i⨂​ρi​∥i⨂​σi​)≤n1​i∑​Di​+n​1​(g(ε)+2nlog2​e∑i​Vi​​)+nC​.

This result records the source’s operational or structural relation, with the stated endpoint and attainment clauses retained.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Intakenext100_P07
Formal statement
noncomputable section

open Filter

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

namespace QAlgorithms.Papers.Independent.Intakenext100.P07

attribute [local instance] Classical.propDecidable Classical.decEq

variable {I J A B C D X Y Z : Type*}

variable [Fintype I] [DecidableEq I] [Fintype J] [DecidableEq J]

variable [Fintype A] [DecidableEq A] [Fintype B] [DecidableEq B]

variable [Fintype C] [DecidableEq C] [Fintype D] [DecidableEq D]

variable [Fintype X] [DecidableEq X] [Fintype Y] [DecidableEq Y]

variable [Fintype Z] [DecidableEq Z]

/-- NODE paper:1504.00233#6.noniid
§6.4.2 (6.97)–(6.100): product-state finite-size bound, arXiv:1504.00233v5, PDF p.109. -/
theorem product_finite_size_bound [Nonempty I] (S : Finset (paperMat I × paperMat I))
    (hS : ∀ x ∈ S, IsDensityMatrix x.1 ∧ IsDensityMatrix x.2 ∧ paperSupported x.1 x.2) :
    ∃ C : ℝ, 0 ≤ C ∧ ∀ n : ℕ, 0 < n → ∀ r s : Fin n → paperMat I, (∀ i, (r i,s i) ∈ S) →
      ∀ e : ℝ, 0 < e → e < 1 →
      ((1/(n:ℝ):ℝ):EReal)*smoothMaxD (productState r) (productState s) e ≤ (((∑ i, (paperRelativeD (r i) (s i)).toReal)/(n:ℝ)+
          (paperPenalty e+(∑ i, paperVariance (r i) (s i))/(2*Real.logb 2 (Real.exp 1)*(n:ℝ)))/Real.sqrt n+C/(n:ℝ):ℝ):EReal) := by sorry

end QAlgorithms.Papers.Independent.Intakenext100.P07

end
Source
Quantum Information Processing with Finite Resources -- Mathematical Foundations, arXiv:1504.00233v5, PDF p. 109, §6.4.2 (6.97)–(6.100): product-state finite-size 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 SSS be a finite set of pairs of matrices on III such that every (r,s)∈S(r,s)\in S(r,s)∈S consists of positive semidefinite matrices of complex trace 111 satisfying ker⁡s⊆ker⁡r\ker s\subseteq\ker rkers⊆kerr. For positive semidefinite matrices define L(X)=(t↦log⁡2t)(X)L(X)=(t\mapsto\log_2 t)(X)L(X)=(t↦log2​t)(X) by functional calculus, using log⁡20=0\log_2 0=0log2​0=0, and set

D(X∥Y)={Re⁡tr⁡(X(L(X)−L(Y)))Re⁡tr⁡Xker⁡Y⊆ker⁡X,+∞otherwise,D(X\|Y)= \begin{cases} \displaystyle \frac{\operatorname{Re}\operatorname{tr}(X(L(X)-L(Y)))} {\operatorname{Re}\operatorname{tr}X}&\ker Y\subseteq\ker X,\\ +\infty&\text{otherwise}, \end{cases}D(X∥Y)=⎩⎨⎧​RetrXRetr(X(L(X)−L(Y)))​+∞​kerY⊆kerX,otherwise,​

with the first branch embedded into the extended reals, and

V(X∥Y)=Re⁡tr⁡ ⁣[X(L(X)−L(Y)−D(X∥Y)R 1)2],V(X\|Y)= \operatorname{Re}\operatorname{tr}\!\left[ X\left(L(X)-L(Y)-D(X\|Y)_{\mathbb R}\,1\right)^2 \right],V(X∥Y)=Retr[X(L(X)−L(Y)−D(X∥Y)R​1)2],

where hRh_{\mathbb R}hR​ is the extended-real-to-real conversion, equal to 000 at either infinity. For subnormalized positive semidefinite X,YX,YX,Y, put

P(X,Y)=1−(Re⁡tr⁡X YX+(1−Re⁡tr⁡X)(1−Re⁡tr⁡Y))2,P(X,Y)= \sqrt{1-\left( \operatorname{Re}\operatorname{tr}\sqrt{\sqrt X\,Y\sqrt X} +\sqrt{(1-\operatorname{Re}\operatorname{tr}X)(1-\operatorname{Re}\operatorname{tr}Y)} \right)^2},P(X,Y)=1−(RetrX​YX​​+(1−RetrX)(1−RetrY)​)2​,

and define

Dmax⁡(X∥Y)=inf⁡{ℓ∈R:X⪯2ℓY},Dmax⁡ e(X∥Y)=inf⁡T⪰0, Re⁡tr⁡T≤1P(T,X)≤eDmax⁡(T∥Y),D_{\max}(X\|Y)= \inf\{\ell\in\mathbb R:X\preceq2^\ell Y\}, \qquad D_{\max}^{\,e}(X\|Y)= \inf_{\substack{T\succeq0,\ \operatorname{Re}\operatorname{tr}T\le1\\P(T,X)\le e}} D_{\max}(T\|Y),Dmax​(X∥Y)=inf{ℓ∈R:X⪯2ℓY},Dmaxe​(X∥Y)=T⪰0, RetrT≤1P(T,X)≤e​inf​Dmax​(T∥Y),

with both infima taken in the extended reals and X⪯YX\preceq YX⪯Y meaning Y−X⪰0Y-X\succeq0Y−X⪰0. There exists a real constant C≥0C\ge0C≥0 such that, for every integer n>0n>0n>0, every two sequences ri,sir_i,s_iri​,si​ indexed by i∈Fin⁡(n)i\in\operatorname{Fin}(n)i∈Fin(n) with (ri,si)∈S(r_i,s_i)\in S(ri​,si​)∈S, and every real 0<e<10<e<10<e<1, the product matrices

Rxy=∏i(ri)xiyi,Txy=∏i(si)xiyiR_{xy}=\prod_i(r_i)_{x_i y_i},\qquad T_{xy}=\prod_i(s_i)_{x_i y_i}Rxy​=i∏​(ri​)xi​yi​​,Txy​=i∏​(si​)xi​yi​​

satisfy

1nDmax⁡ e(R∥T)≤∑iD(ri∥si)Rn+−log⁡2(1−1−e2)+∑iV(ri∥si)2log⁡2(exp⁡1) nn+Cn,\frac1nD_{\max}^{\,e}(R\|T) \le \frac{\sum_iD(r_i\|s_i)_{\mathbb R}}n + \frac{ -\log_2(1-\sqrt{1-e^2}) +\displaystyle\frac{\sum_iV(r_i\|s_i)} {2\log_2(\exp 1)\,n} }{\sqrt n} +\frac Cn,n1​Dmaxe​(R∥T)≤n∑i​D(ri​∥si​)R​​+n​−log2​(1−1−e2​)+2log2​(exp1)n∑i​V(ri​∥si​)​​+nC​,

where the right side is embedded into the extended reals. The constant is chosen before nnn, the sequences, and eee; n=0n=0n=0 and e=0,1e=0,1e=0,1 are excluded. The set SSS may be empty, in which case no sequences satisfy the membership hypothesis for n>0n>0n>0.

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