An uncorrelated unknown qubit costs one qubit
OpenQAlgorithms.Papers.Independent.Gap20.P11.exampleUnknownQubitAlice holds a maximally mixed qubit and Bob has a trivial register, so . The state may be purified by an inaccessible reference. With entropy in bits,
One qubit suffices by direct transmission: the explicit code has a single two-dimensional quantum message and its encoding and decoding compose to the identity on every qubit operator. Alternatively, Alice can distribute one Bell pair by sending one of its halves and then teleport her qubit. The four normalized Bell-measurement branches form a complete instrument, and Bob's two-bit-conditioned corrections are channels. With that one supplied Bell pair, the teleportation protocol acts as the identity on every input operator, including every off-diagonal block of an arbitrary external reference.
The lower bound concerns unknown quantum information with its reference correlations, while the two constructions demonstrate the source's direct and teleportation alternatives.
import Definitions.Def_QAlgorithms_Papers_Independent_Gap20_P11
noncomputable section
namespace QAlgorithms.Papers.Independent.Gap20.P11
open scoped Matrix ComplexConjugate
open QAlgorithms.Papers.Independent.Gap48.P20 (Copies)
open QAlgorithms.Papers.Independent.Information100.P67 (conditionalEntropy)
attribute [local instance] Classical.propDecidable Classical.decEq
/-- PDF pp.3–4, example 1: one qubit is necessary asymptotically, and both
direct transmission and teleportation using one distributed EPR pair work.
Teleportation is asserted on all operator blocks, hence also on an unknown
qubit entangled with an inaccessible reference. -/
theorem exampleUnknownQubit :
IsDensityMatrix unknownQubit ∧ conditionalEntropy unknownQubit = 1 ∧
(∀ rate : ℝ, MergingAchievable unknownQubit rate ↔ 1 ≤ rate) ∧
directQubit.q = 1 ∧ IsChannel directQubit.encode ∧ IsChannel directQubit.decode ∧
(∀ X : Matrix Bool Bool ℂ, directQubit.decode (directQubit.encode X) = X) ∧
IsState (QAlgorithms.Papers.Independent.Information100.P22.bell Bool) ∧ IsKrausFamily teleportRow ∧
(∀ s, IsChannel (conjSuperop (teleportCorrection s))) ∧
(∀ (R : Type) [Fintype R] (X : Matrix (R × Bool) (R × Bool) ℂ),
idTensorSuperop teleportOperation (teleportReferenceInput X) =
fun p q : R × (Unit × Bool) => X (p.1, p.2.2) (q.1, q.2.2)) := by sorry
end QAlgorithms.Papers.Independent.Gap20.P11
end
Read-back
What the Lean code literally says, in plain math · gpt-6.1-sol
Identify the Boolean values with , write for the corresponding standard basis vector, and put . For a matrix , write , and write , with these matrix functions interpreted by continuous functional calculus. The matrix on given by is positive semidefinite with trace , and its conditional entropy equals . For every real rate , the following condition holds if and only if : for every and , there is a natural number such that for every natural with , there are natural numbers and an operation from Alice’s registers and ancillary qubits together with Bob’s registers and ancillary qubits to Alice’s ancillary qubits together with Bob’s registers and ancillary qubits, satisfying , such that , after reindexing all local bases by finite ordinal sets, is generated recursively by identity operations and finite local Kraus families satisfying , with a recursively generated continuation for each outcome and summation over outcomes, and such that the following reconstruction condition holds for every unit vector on , where , whose partial trace over its first factor is : applying to the nonreference registers of gives a density matrix within trace norm of the density matrix of , with Alice’s original Boolean registers moved to Bob and the reference unchanged; here , including . The direct encoding uses one qubit: its encoding is for one-bit strings , and its decoding is , where is the one-bit string with value ; both maps are channels, meaning that each admits a finite Kraus representation with , and for every complex matrix . The vector has norm . For each pair , define the row and matrix by when , and otherwise, and when , and otherwise. Then , and every map is a channel in the preceding finite Kraus sense. Finally, for every finite type , including the empty type, and every complex matrix on , form the matrix with entries on . Applying the identity on and the operation produces the matrix whose entry at is ; this identity is asserted for all , without positivity or normalization assumptions.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.