Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Corollary 20: the infimum over all generalized representations

Open
QAlgorithms.Papers.Independent.Intakenext100.P36.cb_factorization_infimum

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

prove2quantum

Let n,k≥1n,k\ge1n,k≥1 and let ϕ:Mn→Mk\phi:M_n\to M_kϕ:Mn​→Mk​ be complex-linear. Let Rϕ\mathcal R_\phiRϕ​ consist of every finite pair of families Ai∈Mk,nA_i\in M_{k,n}Ai​∈Mk,n​, Bi∈Mn,kB_i\in M_{n,k}Bi​∈Mn,k​ representing ϕ(X)=∑iAiXBi\phi(X)=\sum_iA_iXB_iϕ(X)=∑i​Ai​XBi​ for every matrix XXX. This collection is nonempty, and

∥ϕ∥cb=∥ϕ†∥⋄=inf⁡(A,B)∈Rϕ∥∑iAiAi†∥∥∑iBi†Bi∥.\|\phi\|_{\rm cb}=\|\phi^\dagger\|_\diamond=\inf_{(A,B)\in\mathcal R_\phi}\sqrt{\left\|\sum_i A_iA_i^\dagger\right\|}\sqrt{\left\|\sum_i B_i^\dagger B_i\right\|}.∥ϕ∥cb​=∥ϕ†∥⋄​=(A,B)∈Rϕ​inf​​i∑​Ai​Ai†​​​​i∑​Bi†​Bi​​​.

The infimum allows arbitrary finite lengths, including the empty zero-map representation. The dual is characterized by Tr⁡(ϕ(X)Y)=Tr⁡(Xϕ†(Y))\operatorname{Tr}(\phi(X)Y)=\operatorname{Tr}(X\phi^\dagger(Y))Tr(ϕ(X)Y)=Tr(Xϕ†(Y)). Matrix norms are Euclidean operator norms, and trace norm is unnormalized.

The result turns the stabilized norm into an optimization over actual finite factorizations. Attainment with at most nknknk terms is retained separately in Theorem 19.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Intakenext100_P36
Formal statement
noncomputable section

open scoped Matrix ComplexConjugate ENNReal ComplexOrder

namespace QAlgorithms.Papers.Independent.Intakenext100.P36

local notation "Mat" => QAlgorithms.Papers.Independent.Information100.P50.M

local notation "Map" => QAlgorithms.Papers.Independent.Information100.P50.Lin

open QAlgorithms.Papers.Independent.Dependency50.P49 (traceDual)

/-- Corollary 20, PDF p.8: all finite generalized representations, without a
length cap in the infimum; nonemptiness is asserted rather than assumed. -/
theorem cb_factorization_infimum {n k : ℕ} (hn : 0 < n) (hk : 0 < k)
    (phi : Map n k) :
    (representationCosts phi).Nonempty ∧
    extendedCBNorm phi = rectangularDiamondNorm (traceDual phi) ∧
    extendedCBNorm phi = sInf (representationCosts phi) := by sorry

end QAlgorithms.Papers.Independent.Intakenext100.P36

end
Source
Computing Stabilized Norms for Quantum Operations via the Theory of Completely Bounded Maps, arXiv:0711.3636v1, PDF p. 8, Corollary 20
Read-back

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

Let n,kn,kn,k be positive natural numbers and let ϕ:Mn(C)→Mk(C)\phi:M_n(\mathbb C)\to M_k(\mathbb C)ϕ:Mn​(C)→Mk​(C) be any complex linear map. Let R(ϕ)⊆[0,∞]\mathcal R(\phi)\subseteq[0,\infty]R(ϕ)⊆[0,∞] consist of all numbers

[∥∑i<mAiAi†∥op ∥∑i<mBi†Bi∥op]+\left[ \sqrt{\left\|\sum_{i<m}A_iA_i^\dagger\right\|_{\mathrm{op}}}\, \sqrt{\left\|\sum_{i<m}B_i^\dagger B_i\right\|_{\mathrm{op}}} \right]_+​​i<m∑​Ai​Ai†​​op​​​i<m∑​Bi†​Bi​​op​​​+​

obtained from any natural number mmm and matrices Ai∈Ck×nA_i\in\mathbb C^{k\times n}Ai​∈Ck×n, Bi∈Cn×kB_i\in\mathbb C^{n\times k}Bi​∈Cn×k satisfying ϕ(X)=∑i<mAiXBi\phi(X)=\sum_{i<m}A_iXB_iϕ(X)=∑i<m​Ai​XBi​ for every XXX. Here [t]+[t]_+[t]+​ sends ttt to max⁡(t,0)\max(t,0)max(t,0) in [0,∞][0,\infty][0,∞], and †\dagger† denotes conjugate transpose. Define

C(ψ)=[sup⁡q∈N, X∈Mqu(C)∥X∥op≤1∥(idq⊗ψ)(X)∥op]+,D(ψ)=sup⁡X∈Mu2(C)∥X∥1≤1[∥(idu⊗ψ)(X)∥1]+C(\psi)= \left[\sup_{\substack{q\in\mathbb N,\ X\in M_{qu}(\mathbb C)\\\|X\|_{\mathrm{op}}\le1}} \|(\mathrm{id}_q\otimes\psi)(X)\|_{\mathrm{op}}\right]_+, \qquad D(\psi)= \sup_{\substack{X\in M_{u^2}(\mathbb C)\\\|X\|_1\le1}} [\|(\mathrm{id}_u\otimes\psi)(X)\|_1]_+C(ψ)=​q∈N, X∈Mqu​(C)∥X∥op​≤1​sup​∥(idq​⊗ψ)(X)∥op​​+​,D(ψ)=X∈Mu2​(C)∥X∥1​≤1​sup​[∥(idu​⊗ψ)(X)∥1​]+​

for ψ:Mu(C)→Mv(C)\psi:M_u(\mathbb C)\to M_v(\mathbb C)ψ:Mu​(C)→Mv​(C), where the supremum defining CCC is real, amplification acts on matrix blocks, and ∥X∥1=Re⁡tr⁡X†X\|X\|_1=\operatorname{Re}\operatorname{tr}\sqrt{X^\dagger X}∥X∥1​=RetrX†X​. If ϕ∨\phi^\veeϕ∨ is the trace dual defined by tr⁡(ϕ(X)Y)=tr⁡(Xϕ∨(Y))\operatorname{tr}(\phi(X)Y)=\operatorname{tr}(X\phi^\vee(Y))tr(ϕ(X)Y)=tr(Xϕ∨(Y)), then

R(ϕ)≠∅,C(ϕ)=D(ϕ∨),C(ϕ)=inf⁡R(ϕ).\mathcal R(\phi)\ne\varnothing, \qquad C(\phi)=D(\phi^\vee), \qquad C(\phi)=\inf\mathcal R(\phi).R(ϕ)=∅,C(ϕ)=D(ϕ∨),C(ϕ)=infR(ϕ).

The infimum is in [0,∞][0,\infty][0,∞]; this statement does not assert that it is attained. Empty representations m=0m=0m=0 are allowed and represent the zero map.

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