A band family for the trial construction: m pairwise disjoint
measurable subsets of (0, ∞) with positive band values. At
consumption these are the retained shifted bins I_j^{θ₀} ∩ [L, H]
and their upper endpoints t_j (06_otqcs.tex, eq N1-Z "the finite
retained bin set").
- meas (j : Fin m) : MeasurableSet (B j)
- disj : Pairwise (Function.onFun Disjoint B)
Instances For
Z = ∑_{j ∈ J} t_j² (06_otqcs.tex, eq N1-Z).
Equations
- CommutingRepetition.bandZ t = ∑ j : Fin m, t j ^ 2
Instances For
Alice's band mass a_s = ∑_j t_j² τ(p_j(h_s)) in finite band
form (06_otqcs.tex, eq abcGamma via eq one-trial-probabilities).
Equations
Instances For
Bob's band mass b_t = ∑_j t_j² τ(p_j(k_t)).
Equations
Instances For
The joint band mass c_{st} = ∑_j t_j² τ(p_j(h_s) p_j(k_t)), read
through the joint spectral coupling (06_otqcs.tex, eqs abcGamma +
joint-measure).
Equations
Instances For
Core w-norm computation (06_otqcs.tex, eq w-norm):
τ(w* w) = ∑_j t_j² τ(p_j(h) p_j(k)) = c, where
w = ∑_j t_j p_j(h) p_j(k) v.
The one-trial resource matrix: √((m+1)/Z) ∑_{j} t_j e_jj ⊗ 1,
zero at the distinguished diagonal entry ★★ (06_otqcs.tex, eq
omega0, as an element of M_{m+1}(N) before the GNS embedding).
Equations
- CommutingRepetition.trialUnitMat N m t = Matrix.diagonal fun (i : Fin (m + 1)) => Fin.lastCases 0 (fun (j : Fin m) => ↑(√((↑m + 1) / CommutingRepetition.bandZ t) * t j) • 1) i
Instances For
The one-trial resource vector ω₀ ∈ L²(M_{m+1}(N))
(06_otqcs.tex, eq omega0).
Equations
- CommutingRepetition.trialState N m t = (N.amplify (m + 1)).ι (CommutingRepetition.trialUnitMat N m t)
Instances For
Alice's one-trial success projection
P_s = ∑_j e_jj ⊗ p_j(h_s) (06_otqcs.tex, eq PQ).
Equations
- CommutingRepetition.trialPA N F B s = Matrix.diagonal fun (i : Fin (m + 1)) => Fin.lastCases 0 (fun (j : Fin m) => (F.dataA s).proj (B j)) i
Instances For
Bob's one-trial success projection
Q_t = ∑_j e_jj ⊗ p_j(k_t) (06_otqcs.tex, eq PQ).
Equations
- CommutingRepetition.trialQB N F B t' = Matrix.diagonal fun (i : Fin (m + 1)) => Fin.lastCases 0 (fun (j : Fin m) => (F.dataB t').proj (B j)) i
Instances For
Alice's bin-erasing partial isometry
V_s = ∑_j e_{★j} ⊗ p_j(h_s) (06_otqcs.tex, eq erasers).
Equations
Instances For
Bob's bin-erasing partial isometry
W_t = ∑_j e_{j★} ⊗ p_j(k_t) (06_otqcs.tex, eq erasers).
Equations
Instances For
The polar corrector v̄_t = e_{★★} ⊗ v_t (06_otqcs.tex, above eq
polar-initial).
Equations
- CommutingRepetition.trialVbar N F t' = Matrix.single (Fin.last m) (Fin.last m) (F.dataB t').v
Instances For
The erased, polar-corrected common branch
ζ_{st} = V_s ω₀ W_t v̄_t — Alice's eraser on the left, Bob's on the
right (06_otqcs.tex, eq common-branch).
Equations
- One or more equations did not get rendered due to their size.
Instances For
The one-trial resource is a unit vector (06_otqcs.tex, below eq omega0).
Alice's one-trial success probability is a_s / Z (06_otqcs.tex,
eq one-trial-probabilities, first entry).
Bob's one-trial success probability is b_t / Z (06_otqcs.tex, eq
one-trial-probabilities, second entry).
The joint one-trial success probability is c_{st} / Z
(06_otqcs.tex, eq one-trial-probabilities, third entry).
V_s* V_s = P_s (06_otqcs.tex, below eq erasers).
W_t W_t* = Q_t (06_otqcs.tex, below eq erasers).
W_t v̄_t v̄_t* W_t* = Q_t: every retained spectral projection of
k_t lies below the left support v_t v_t* (06_otqcs.tex, eq
polar-initial).
The erased, polar-corrected common branch as a matrix (eq common-branch before the
GNS map): V_s ω₀ (W_t v̄_t) = e_{★★} ⊗ √((m+1)/Z) w_{st}.
The common branch in closed form:
ζ_{st} = √((m+1)/Z) · ι(e_{★★} ⊗ ∑_j t_j p_j(h_s) p_j(k_t) v_t)
(06_otqcs.tex, eq common-branch, second line).
‖ζ_{st}‖² = c_{st} / Z (06_otqcs.tex, below eq common-branch).
One-trial traces and the corner trace (proof-side helpers for OTQCS/Compile) #
τ₁(ω₀* ω₀) = 1 (the unit-norm resource, in trace form).
τ₁(ω₀* P_s ω₀) = a_s/Z (eq one-trial-probabilities, trace form).
τ₁(ω₀* ω₀ Q_t) = b_t/Z.
τ₁(ω₀* P_s ω₀ Q_t) = c_{st}/Z.
First-success arithmetic (node 1.3.6) #
Scalar consumed form of lem otqcs-first-success: the per-trial success
probabilities are a/Z, b/Z, c/Z (eq one-trial-probabilities), and
independence across trials reduces the first-success analysis to the
three scalar facts below (exact geometric mass of the equal-index
event, the mismatch bound, and the exhaustion tail).
Exact equal-index probability (06_otqcs.tex, proof of lem
otqcs-first-success, the geometric sum): with per-trial progress
w = (a + b − c)/Z, the total mass of "both first successes at the
same trial" is ∑_{j≥0} (1 − w)^j (c/Z) = c/(a + b − c).
First-success mismatch (node 1.3.6; 06_otqcs.tex, eq
first-index-mismatch): when a, b ≥ 1/2 (eq ab-mass under the
standing condition ρ + L² ≤ 1/4) and c ≤ min{a, b} (eq c-min),
the index-mismatch probability (a + b − 2c)/(a + b − c) is at most
2 Γ = 2 (a + b − 2c).