Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

eq37

Open
QAlgorithms.Papers.Independent.Support50.P15.eq37

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

prove2quantum

Let k≥1k\ge1k≥1 and let ρ0,…,ρk\rho_0,\ldots,\rho_kρ0​,…,ρk​ be normalized states in a common positive finite dimension. No common-support assumption is needed. Writing ∥⋅∥1\|\cdot\|_1∥⋅∥1​ for the trace norm,

∑j=1kD(ρj−1∥ρj)≥12∑j=1k∥ρj−1−ρj∥12≥12k(∑j=1k∥ρj−1−ρj∥1)2≥∥ρ0−ρk∥122k.\sum_{j=1}^kD(\rho_{j-1}\|\rho_j)\ge\frac12\sum_{j=1}^k\|\rho_{j-1}-\rho_j\|_1^2\ge\frac{1}{2k}\left(\sum_{j=1}^k\|\rho_{j-1}-\rho_j\|_1\right)^2\ge\frac{\|\rho_0-\rho_k\|_1^2}{2k}.j=1∑k​D(ρj−1​∥ρj​)≥21​j=1∑k​∥ρj−1​−ρj​∥12​≥2k1​(j=1∑k​∥ρj−1​−ρj​∥1​)2≥2k∥ρ0​−ρk​∥12​​.

Support failures give infinity on the left. This retains the full intermediate chain and supplies an endpoint-distance lower bound even when the entropy change vanishes.

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Support50_P15
Formal statement
noncomputable section

attribute [local instance] Classical.propDecidable

open scoped Matrix ComplexConjugate ENNReal Matrix.Norms.L2Operator

open Filter MeasureTheory

namespace QAlgorithms.Papers.Independent.Support50.P15

open QAlgorithms.Papers.Independent.Information100.P56 (entropy)

/-- Eq. (37), PDF p.12, with every intermediate term and no unnecessary full-rank premise. -/
theorem eq37 {d k : ℕ} (hd : 0 < d) (hk : 0 < k) (a : Fin (k + 1) → EntropyState d) :
    let total := ∑ j : Fin k, stateRelativeNats (a j.castSucc) (a j.succ)
    let sq : ℝ := ∑ j : Fin k, (traceNorm ((a j.castSucc).val - (a j.succ).val)) ^ 2 / 2
    let av : ℝ := (k : ℝ) / 2 * (∑ j : Fin k,
      traceNorm ((a j.castSucc).val - (a j.succ).val) / k) ^ 2
    (sq : EReal) ≤ total ∧ av ≤ sq ∧
    (traceNorm ((a 0).val - (a (Fin.last k)).val)) ^ 2 / (2 * (k : ℝ)) ≤ av := by sorry

end QAlgorithms.Papers.Independent.Support50.P15

end
Source
Tight bound on relative entropy by entropy difference, arXiv:1304.0036v3, PDF p. 12, eq37
Read-back

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

For all positive natural numbers d,kd,kd,k and every sequence A0,…,AkA_0,\ldots,A_kA0​,…,Ak​ of positive semidefinite complex d×dd\times dd×d matrices of trace 111, define ∥M∥1=Re⁡Tr⁡M∗M\|M\|_1=\operatorname{Re}\operatorname{Tr}\sqrt{M^*M}∥M∥1​=ReTrM∗M​, using spectral functional calculus. Define D(A∥B)=+∞D(A\Vert B)=+\inftyD(A∥B)=+∞ if ker⁡B⊈ker⁡A\ker B\not\subseteq\ker AkerB⊆kerA, and otherwise

D(A∥B)=max⁡{0,Re⁡Tr⁡[(u↦ulog⁡u)(A)−Alog⁡B]},D(A\Vert B)=\max\left\{0,\operatorname{Re}\operatorname{Tr}\bigl[(u\mapsto u\log u)(A)-A\log B\bigr]\right\},D(A∥B)=max{0,ReTr[(u↦ulogu)(A)−AlogB]},

with log⁡0=0\log0=0log0=0. Put

R=∑j=0k−1D(Aj∥Aj+1),Q=12∑j=0k−1∥Aj−Aj+1∥12,Aav=k2(∑j=0k−1∥Aj−Aj+1∥1k)2.R=\sum_{j=0}^{k-1}D(A_j\Vert A_{j+1}),\qquad Q=\frac12\sum_{j=0}^{k-1}\|A_j-A_{j+1}\|_1^2,\qquad A_{\mathrm{av}}=\frac{k}{2} \left(\sum_{j=0}^{k-1}\frac{\|A_j-A_{j+1}\|_1}{k}\right)^2.R=j=0∑k−1​D(Aj​∥Aj+1​),Q=21​j=0∑k−1​∥Aj​−Aj+1​∥12​,Aav​=2k​(j=0∑k−1​k∥Aj​−Aj+1​∥1​​)2.

Then Q≤RQ\le RQ≤R, where QQQ is embedded into the extended reals, Aav≤QA_{\mathrm{av}}\le QAav​≤Q, and

∥A0−Ak∥122k≤Aav.\frac{\|A_0-A_k\|_1^2}{2k}\le A_{\mathrm{av}}.2k∥A0​−Ak​∥12​​≤Aav​.

No full-rank assumption is imposed, so RRR is allowed to be +∞+\infty+∞; dimension d=1d=1d=1 is included.

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