Definitions for Is Entanglement Monogamous?
DefinitionQAlgorithms_Papers_Independent_Intakethird100_P06The objects the formalized statements of Is Entanglement Monogamous? (arXiv:quant-ph/0307120) are stated in. Each was checked against the paper by a blind read-back and a moderator.
Opinion: PDF p.5: each opinion system is one qubit in the displayed 0/1 basis.eprDensity: PDF p.5, (1): the normalized plus-sign EPR density, using the frozen vector.IsCoordinateDensity: PDF pp.5-7: positive trace-class density operators of trace one. This paper-local adapter reuses the pinned accepted predicate verbatim: nuclearity, real nonnegative finite quadratic forms, summable diagonal and trace one. No global-purity, finite-rank or countable-basis restriction is imposed.swapBobJoe: PDF pp.5,7: interchange Bob and Joe, fixing Alice; both are the same qubit.aliceJoeMarginal: PDF p.7, weak extension terminology applied to the p.5 example: trace Bob out, leaving Alice and Joe in that order.eveBlock: PDF pp.5-6: the unnormalized Eve operator for the outcome (a,b) of computational-basis measurement of both qubits. Zero blocks remain zero; there is no division by the outcome probability.sharedBitLaw: PDF p.5: classical common-bit law on n parties. p is the probability of 0 (false); 1-p is the probability of 1 (true). The two contributions are added, so even at n=0 this raw formula remains a normalized one-point law; the sharing theorem requires at least one actual party.
import Mathlib
import Definitions.Def_QAlgorithms_Defs_Channel
import Definitions.Def_QAlgorithms_Defs_RealLabelRegisters
import Definitions.Def_QAlgorithms_Defs_Measurement
import Definitions.Def_QAlgorithms_Papers_Independent_Dependency50_P33
import Definitions.Def_QAlgorithms_Papers_Independent_Frontier18_P09
/-!
# Is Entanglement Monogamous? Elementary statements
Source: Barbara M. Terhal, arXiv:quant-ph/0307120v1, PDF pp.5-7.
All new definitions are local to this paper. Nuclear expansions and density
positivity are reused from pinned accepted Dependency50.P33, lines 29-60.
The unrestricted partial trace is reused from pinned accepted Frontier18.P09.
These accepted-library imports follow the independent-paper override. Arbitrary orthonormal index types retain
arbitrary Hilbert dimension, including nonseparable spaces. No Countable or
Fintype hypothesis is imposed on the third system.
Equation (2), p.5, prints Joe as a ket. The statements retain the prose's actual
hypothesis (only AB is pure), and use the computed, possibly mixed, third marginal.
This correction of ket shorthand requires independent moderation. All theorem
proofs are intentionally sorry. Compilation is pending the batch handoff.
-/
noncomputable section
namespace QAlgorithms.Papers.Independent.Intakethird100.P06
open scoped BigOperators
open QAlgorithms.Papers.Independent.Dependency50.P33 (NuclearExpansion)
open QAlgorithms.Papers.Independent.Frontier18.P09 (partialTraceRightCoeff)
attribute [local instance] Classical.propDecidable Classical.decEq
/-- PDF p.5: each opinion system is one qubit in the displayed 0/1 basis. -/
abbrev Opinion := Qubits 1
/-- PDF p.5, (1): the normalized plus-sign EPR density, using the frozen vector. -/
def eprDensity : Matrix (Opinion × Opinion) (Opinion × Opinion) ℂ :=
pureDensity (maxEntangled 1)
/-- PDF pp.5-7: positive trace-class density operators of trace one.
This paper-local adapter reuses the pinned accepted predicate verbatim:
nuclearity, real nonnegative finite quadratic forms, summable diagonal and trace
one. No global-purity, finite-rank or countable-basis restriction is imposed. -/
abbrev IsCoordinateDensity {I : Type} (X : Matrix I I ℂ) : Prop :=
QAlgorithms.Papers.Independent.Dependency50.P33.IsDensity X
/-- PDF pp.5,7: interchange Bob and Joe, fixing Alice; both are the same qubit. -/
def swapBobJoe (X : Matrix ((Opinion × Opinion) × Opinion)
((Opinion × Opinion) × Opinion) ℂ) :
Matrix ((Opinion × Opinion) × Opinion) ((Opinion × Opinion) × Opinion) ℂ :=
fun x y => X ((x.1.1, x.2), x.1.2) ((y.1.1, y.2), y.1.2)
/-- PDF p.7, weak extension terminology applied to the p.5 example:
trace Bob out, leaving Alice and Joe in that order. -/
def aliceJoeMarginal (X : Matrix ((Opinion × Opinion) × Opinion)
((Opinion × Opinion) × Opinion) ℂ) : Matrix (Opinion × Opinion) (Opinion × Opinion) ℂ :=
fun x y => ∑ b : Opinion, X ((x.1, b), x.2) ((y.1, b), y.2)
/-- PDF pp.5-6: the unnormalized Eve operator for the outcome (a,b)
of computational-basis measurement of both qubits. Zero blocks remain zero;
there is no division by the outcome probability. -/
def eveBlock {E : Type} (X : Matrix ((Opinion × Opinion) × E)
((Opinion × Opinion) × E) ℂ) (a b : Opinion) : Matrix E E ℂ :=
fun e f => X ((a, b), e) ((a, b), f)
/-- PDF p.5: classical common-bit law on n parties. p is the probability
of 0 (false); 1-p is the probability of 1 (true). The two contributions are
added, so even at n=0 this raw formula remains a normalized one-point law;
the sharing theorem requires at least one actual party. -/
def sharedBitLaw (n : ℕ) (p : ℝ) (x : Fin n → Bool) : ℝ :=
(if x = (fun _ => false) then p else 0) +
(if x = (fun _ => true) then 1 - p else 0)
/-- Nuclearity includes the zero operator, while normalization excludes it as a state. -/
example {I : Type} : Nonempty (NuclearExpansion (0 : Matrix I I ℂ)) := by
refine ⟨{ left := fun _ => 0
right := fun _ => 0
summable := ?_
coefficients := ?_ }⟩
· simp
· intro i j
simp
example {I : Type} : ¬ IsCoordinateDensity (0 : Matrix I I ℂ) := by
intro h
simpa using h.2.2.2
example {A E : Type} :
partialTraceRightCoeff (0 : Matrix (A × E) (A × E) ℂ) = 0 := by
funext a b
simp [partialTraceRightCoeff]
example {A E : Type} [Fintype E] (X : Matrix (A × E) (A × E) ℂ) :
partialTraceRightCoeff X = traceRight X := by
funext a b
simp [partialTraceRightCoeff, traceRight]
/-- The mismatched 01 outcome has zero EPR probability. -/
example : eprDensity ((fun _ : Fin 1 => false), (fun _ => true))
((fun _ : Fin 1 => false), (fun _ => true)) = 0 := by
have h : (fun _ : Fin 1 => false) ≠ (fun _ : Fin 1 => true) := by decide
simp [eprDensity, pureDensity, maxEntangled, h]
example {E : Type} (X : Matrix ((Opinion × Opinion) × E)
((Opinion × Opinion) × E) ℂ) (a b : Opinion) (e f : E) :
eveBlock X a b e f = X ((a, b), e) ((a, b), f) := by
rfl
example (p : ℝ) : sharedBitLaw 1 p (fun _ => false) = p := by
have h : (fun _ : Fin 1 => false) ≠ (fun _ : Fin 1 => true) := by decide
simp [sharedBitLaw, h]
example (p : ℝ) : sharedBitLaw 0 p (fun _ => false) = 1 := by
have h : (fun _ : Fin 0 => false) = (fun _ : Fin 0 => true) := by
funext i
exact Fin.elim0 i
simp [sharedBitLaw, h] <;> ring
/-- The ownership permutation is an involution. -/
example (X : Matrix ((Opinion × Opinion) × Opinion)
((Opinion × Opinion) × Opinion) ℂ) : swapBobJoe (swapBobJoe X) = X := by
rfl
example : aliceJoeMarginal (0 : Matrix ((Opinion × Opinion) × Opinion)
((Opinion × Opinion) × Opinion) ℂ) = 0 := by
funext a b
simp [aliceJoeMarginal]
end QAlgorithms.Papers.Independent.Intakethird100.P06
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
QAlgorithms.Papers.Independent.Intakethird100.P06.IsCoordinateDensity
For any type and any complex matrix indexed by , means that all the following hold: there exist sequences , indexed by , such that is summable and for every ; for every finite subset and every function , the number has imaginary part zero and nonnegative real part; and the real diagonal family is summable with sum . There is no finiteness or countability assumption on ; when is empty, the diagonal-sum requirement cannot hold.
QAlgorithms.Papers.Independent.Intakethird100.P06.Opinion
is the type of Boolean functions on the one-element set , equivalently the set of one-bit strings. It has two elements, denoted and , whose values at are respectively false and true.
QAlgorithms.Papers.Independent.Intakethird100.P06.aliceJoeMarginal
For every complex matrix indexed by triples of one-bit strings, is the matrix on pairs of one-bit strings whose entries are . The sum runs over the two one-bit strings and uses the same second-coordinate value in the row and column.
QAlgorithms.Papers.Independent.Intakethird100.P06.eprDensity
On pairs of one-bit strings, is the complex matrix with entries , where when and otherwise. Thus when both and , and it is zero otherwise; equality between and is not required.
QAlgorithms.Papers.Independent.Intakethird100.P06.eveBlock
For every type , every complex matrix indexed by with one-bit strings and , and every fixed pair of one-bit strings , is the matrix indexed by whose entries are . No finiteness or countability assumption is imposed on .
QAlgorithms.Papers.Independent.Intakethird100.P06.sharedBitLaw
For every natural number , every real number , and every Boolean string of length , is . For , this gives value at the all-false string, value at the all-true string, and zero elsewhere. For , the two constant strings coincide with the unique empty string, whose value is therefore . The definition places no restriction on .
QAlgorithms.Papers.Independent.Intakethird100.P06.swapBobJoe
For every complex matrix indexed by triples of one-bit strings, is the matrix obtained by interchanging the second and third coordinates in both indices: for all one-bit strings .
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.