Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Signed EPR/coherent expressions: comparison and pure-loss equality

Proved
QAlgorithms.Papers.Independent.Intakenext100.P72.eprCoherentComparison

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

prove2quantum

Let V≥1V\ge1V≥1, G>0G>0G>0, and Gχ≥∣1−G∣G\chi\ge|1-G|Gχ≥∣1−G∣. Write P(G,χ,V,s)=(Gχ+Gs)(Gχ+G/V)P(G,\chi,V,s)=(G\chi+Gs)(G\chi+G/V)P(G,χ,V,s)=(Gχ+Gs)(Gχ+G/V) and Ds=12log⁡2[1/P(G,χ,V,s)]D_s=\tfrac12\log_2[1/P(G,\chi,V,s)]Ds​=21​log2​[1/P(G,χ,V,s)]. The EPR expression includes the probability 1/21/21/2 of matching bases, while the coherent expression uses either basis:

E=12D1/V=14log⁡21(Gχ+G/V)2=12log⁡21Gχ+G/V,C=D1=12log⁡21(Gχ+G/V)(Gχ+G)=E−12log⁡2[G(1+χ)]≤E.\begin{aligned} E&=\tfrac12D_{1/V}=\frac14\log_2\frac1{(G\chi+G/V)^2} =\frac12\log_2\frac1{G\chi+G/V},\\ C&=D_1=\frac12\log_2\frac1{(G\chi+G/V)(G\chi+G)} =E-\frac12\log_2[G(1+\chi)]\le E. \end{aligned}EC​=21​D1/V​=41​log2​(Gχ+G/V)21​=21​log2​Gχ+G/V1​,=D1​=21​log2​(Gχ+G/V)(Gχ+G)1​=E−21​log2​[G(1+χ)]≤E.​

Equality holds precisely when G≤1G\le1G≤1 and Gχ=1−GG\chi=1-GGχ=1−G. On that pure-loss branch,

C=E=12log⁡211−G(1−1/V)≥0.C=E=\frac12\log_2\frac1{1-G(1-1/V)}\ge0.C=E=21​log2​1−G(1−1/V)1​≥0.

The comparison includes physical amplification. Equality concerns signed expressions before clipping at zero; at V=1V=1V=1 the pure-loss expression is zero. Identifying these expressions with measured information is the separate deferred node (45)–(48).

Preamble
import Definitions.Def_QAlgorithms_Papers_Independent_Intakenext100_P72
Formal statement
noncomputable section

open Filter

open scoped Topology

namespace QAlgorithms.Papers.Independent.Intakenext100.P72

/-- Grosshans--Grangier, PDF p.5, IV.A, (54)--(58).
The denominator in (54), inspected in the PDF, is (G*chi+G/V)^2.
Comparison and equality concern signed model expressions before clipping;
physical amplifiers are included, with pure loss confined to G≤1. -/
theorem eprCoherentComparison
    (G chi V : ℝ) (hG : 0 < G) (hV : 1 ≤ V)
    (hphysical : |1 - G| ≤ G * chi) :
    eprExpression G chi V = (1 / 4 : ℝ) *
      Real.logb 2 (1 / (G * chi + G / V) ^ 2) ∧
    eprExpression G chi V = (1 / 2 : ℝ) * Real.logb 2 (1 / (G * chi + G / V)) ∧
    coherentExpression G chi V = (1 / 2 : ℝ) *
      Real.logb 2 (1 / ((G * chi + G / V) * (G * chi + G))) ∧
    coherentExpression G chi V = eprExpression G chi V -
      (1 / 2 : ℝ) * Real.logb 2 (G * (1 + chi)) ∧
    coherentExpression G chi V ≤ eprExpression G chi V ∧
    (coherentExpression G chi V = eprExpression G chi V ↔
      G ≤ 1 ∧ G * chi = 1 - G) ∧
    (G ≤ 1 → G * chi = 1 - G →
      eprExpression G chi V = lossExpression G V ∧
      coherentExpression G chi V = lossExpression G V ∧
      0 ≤ lossExpression G V) := by sorry

end QAlgorithms.Papers.Independent.Intakenext100.P72

end
Source
Reverse reconciliation protocols for quantum cryptography with continuous variables, arXiv:quant-ph/0204127v1, PDF p. 5, IV.A, (54)–(58): EPR/coherent information rates and equality case
Read-back

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

For every G,χ,V∈RG,\chi,V\in\mathbb RG,χ,V∈R satisfying G>0G>0G>0, V≥1V\ge1V≥1, and ∣1−G∣≤Gχ|1-G|\le G\chi∣1−G∣≤Gχ, define

E=14log⁡2 ⁣(1(Gχ+G/V)2),C=12log⁡2 ⁣(1(Gχ+G)(Gχ+G/V)),L=12log⁡2 ⁣(11−G(1−1/V)).E=\frac14\log_2\!\left(\frac1{(G\chi+G/V)^2}\right),\qquad C=\frac12\log_2\!\left(\frac1{(G\chi+G)(G\chi+G/V)}\right),\qquad L=\frac12\log_2\!\left(\frac1{1-G(1-1/V)}\right).E=41​log2​((Gχ+G/V)21​),C=21​log2​((Gχ+G)(Gχ+G/V)1​),L=21​log2​(1−G(1−1/V)1​).

Then all of the following hold:

E=14log⁡2 ⁣(1(Gχ+G/V)2),E=12log⁡2 ⁣(1Gχ+G/V),C=12log⁡2 ⁣(1(Gχ+G/V)(Gχ+G)),C=E−12log⁡2 ⁣(G(1+χ)),C≤E,C=E⟺(G≤1 and Gχ=1−G),(G≤1 and Gχ=1−G)⟹(E=L, C=L, and L≥0).\begin{gathered} E=\frac14\log_2\!\left(\frac1{(G\chi+G/V)^2}\right),\\ E=\frac12\log_2\!\left(\frac1{G\chi+G/V}\right),\\ C=\frac12\log_2\!\left(\frac1{(G\chi+G/V)(G\chi+G)}\right),\\ C=E-\frac12\log_2\!\bigl(G(1+\chi)\bigr),\qquad C\le E,\\ C=E\quad\Longleftrightarrow\quad \bigl(G\le1\ \text{and}\ G\chi=1-G\bigr),\\ \bigl(G\le1\ \text{and}\ G\chi=1-G\bigr) \quad\Longrightarrow\quad \bigl(E=L,\ C=L,\ \text{and}\ L\ge0\bigr). \end{gathered}E=41​log2​((Gχ+G/V)21​),E=21​log2​(Gχ+G/V1​),C=21​log2​((Gχ+G/V)(Gχ+G)1​),C=E−21​log2​(G(1+χ)),C≤E,C=E⟺(G≤1 and Gχ=1−G),(G≤1 and Gχ=1−G)⟹(E=L, C=L, and L≥0).​

The hypotheses include V=1V=1V=1 and impose no upper bound on GGG. All logarithms here are the real base-two logarithm used in the statement, with total real division used in their arguments.

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