Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Definitions for Performance and structure of single-mode bosonic codes

Definition
QAlgorithms_Papers_Independent_Intakenext100_P65

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

prove2quantum

The objects the formalized statements of Performance and structure of single-mode bosonic codes (arXiv:1708.05010) are stated in. Each was checked against the paper by a blind read-back and a moderator.

  • FockCoordinates: PDF pp.2--3: number-basis coordinates, prior to imposing the Hilbert domain.
  • fockDomain: PDF p.2: the full oscillator Hilbert domain, expressed in its countable ONB.
  • fockNormSq: PDF p.2: squared Hilbert norm on the square-summable domain.
  • fockInner: PDF p.2: inner product, conjugate-linear in the first entry.
  • fockMomentDomain: PDF pp.2,4: a finite photon-number moment has a summability domain.
  • numberMoment: PDF pp.4,18: kth number moment, only used on fockMomentDomain.
  • codeMoment: PDF p.4 (2.1): literal half-sum of the two logical number moments.
  • annPower: PDF p.2: coefficient of a^k, including its factorial normalization.
  • annDomain: PDF p.2: natural domain of the unbounded annihilation power.
  • lossCoefficient: PDF p.2 (1.4): the exact loss coefficient for input occupation n, zero below l.
  • lossVector: PDF p.2 (1.4): loss Kraus action on the full number basis.
  • qecBlock: PDF p.8 (4.1): actual compression of E_l^dagger E_k, not a free QEC matrix.
  • qecC: PDF p.8 (4.4): coefficient of the identity.
  • qecZ: PDF p.8 (4.4): coefficient of logical Z in the chosen basis.
  • qecX: PDF p.8 (4.4): coefficient of logical X, including nonsymmetric off-diagonal blocks.
  • liouvilleBasis: PDF p.3 (1.8): I,X,Y,Z, using the accepted Bool-indexed Pauli matrices with Fin 2 labels 0,1 mapped to false,true.
  • liouville: PDF p.3 (1.8): Pauli Liouville matrix; this is distinct from the Choi matrix.
  • traceReference: PDF p.7 (3.3): trace over the reference, leaving the channel's output marginal.
  • reverseInformation: PDF p.7 (3.3): reverse coherentCoordinates information, without capacity optimization. The frozen output-first Choi operator is unnormalized; multiplying by 1/2 gives the trace-one qubit Choi state in (1.9).
  • qubitFrobeniusSq: PDF p.7 (3.4): squared Frobenius norm from accepted Advance50.P29, specialized to the effective qubit channel.
  • worstPureFidelity: PDF p.24 (A2): minimum over normalized pure inputs; compactness and CPTP supply a nonempty bounded set, with attainment asserted below.
  • residueProject: PDF p.9 (5.3): periodic residue projector, allowing signed residue labels.
  • coherentCoordinates: PDF p.9: normalized coherentCoordinates-state coefficients for real amplitude a.
  • catNorm: PDF p.9 (5.2): the actual projected coherentCoordinates-state probability.
  • catCodeCoordinates: PDF p.9 (5.1): normalized projected coherentCoordinates coordinates; a>0 in every use.
  • residueCheck: PDF p.9 (5.7): bounded number-phase rotation.
  • residueLogicalZ: PDF p.9 following (5.7): the specified square root of the check is logical Z.
  • catNormFourier: PDF p.9 (5.5): Fourier projection using exact coherentCoordinates-state overlaps.
  • catLossRatio: PDF pp.9--11: probability ratios with signed residues, before taking any limit.
  • poissonWeight: PDF p.10 (5.12): Poisson weight whose mean g*a^2 varies with a.
  • antipodal: PDF p.14 (6.11): finite spin coefficients, indexed by twice the spin.
  • binomialCodeCoordinates: PDF p.11 (6.1): direct full-Fock binomialCodeCoordinates coordinates.
  • binomialParity: PDF p.12 (6.4): the positive/negative logical combinations.
  • binomialGenerator: PDF p.12 (6.5): exact generating polynomial, with plus/minus sign.
  • spinX: PDF p.14: standard spin-x matrix, on m=0,...,M for spin M/2.
  • embedOne: PDF pp.14--15: transport of a coefficient vector to one mode.
  • embedTwo: PDF p.15 Fig.7(a), p.27: two-mode transport with the spacing in BOTH occupations. Printed (6.16) drops a spacing factor; use the explicit preceding basis map.
  • qubitDickeCoordinates: PDF p.15: accepted Information100.P22.dicke in Fin 2 word coordinates; Bool false,true correspond to 0,1. Infeasible excitation weights give zero.
  • embedDicke: PDF p.15 (6.14): the spaced Dicke transport (K >= M(S+1) at use sites).
  • smallGenerator: PDF p.17: even/odd spin generator on the stated seven-dimensional invariant space.
  • smallUnitary: PDF p.17 (6.21): a=8,b=5, with identity extension outside the invariant space.
  • smallBasis: PDF p.17 Tables II--III: actual normalized code, dephasing residual, loss, and failure vectors. Residual normalization is sqrt(3), correcting the false coefficient 2 in (6.20).
  • totalQuditExcitation: PDF p.28 (C4): totalQuditExcitation of an M-qudit computational word.
  • extendedCount: PDF p.27 (C2),(C4): number of words of a given total totalQuditExcitation.
  • extendedVector: PDF p.27 (C4): normalized extended binomialCodeCoordinates vector, distinct from a general qudit Dicke vector.
  • quditFourierFactor: PDF p.28 (C8): positive Fourier phase, matching the explicit backwards shift (C7).
  • extendedCoefficients: PDF p.27 (C3): totalQuditExcitation-basis coefficients, including the endpoint (d-1)M.
  • permCode: PDF p.27 (C3): the new multi-qudit code, in extended-binomialCodeCoordinates coordinates.
  • fourierProduct: PDF p.28 (C9): tensor product of M identical Fourier vectors.
  • nextLabel: PDF p.28 (C7): the input label one step above the output label.
  • collectiveShift: PDF p.28 (C10): average of the backwards cyclic shifts on individual qudits.
  • compressedShift: PDF p.28: compression to the extended-binomialCodeCoordinates span, of dimension (d-1)M+1.
  • weylPhase: PDF p.28 (D1b): the positive Weyl convention.
  • lossEntry: PDF p.2 (1.4): number-basis kernel of a single loss Kraus operator.
  • lossProduct: PDF p.19 (7.12): genuine product kernel, before the displacement expansion.
  • lossDisplacementIntegrand: PDF p.19 (7.12): integrand INCLUDING the factor 1/pi, on complex-plane Lebesgue measure.
  • fockDensity: PDF pp.2,20: positive trace-one full Fock kernels. Positivity of every finite principal compression and summability of the nonnegative diagonal characterize positive trace-class operators; there is no fixed cutoff.
  • lossChannel: PDF p.2: pure loss, by the countable Kraus sum on actual Fock kernels.
  • amplifierChannel: PDF p.20 (7.20): amplification after loss; Kraus operators sqrt(1-g) E_l^dagger.
  • displaceDensity: PDF p.20 (7.21): displacement conjugation, evaluated on all matrix entries.
  • amplifiedLossDensity: PDF p.20 (7.21): density of the amplified-loss displacement distribution on C.
  • idealComb: PDF p.18 (7.5): distributional square-code comb acting linearly on tests.
  • regularizedLattice: PDF pp.18,29 (7.6),(D5): Gaussian-Abel regularized coherentCoordinates-lattice sum. The ordered product is D_real D_imag; it is evaluated on the vacuum in position coordinates.
  • kernelLattice: PDF p.29 (D9): normalizer-lattice displacement, with signed mu-nu offset.
  • kernelPhase: PDF p.29 (D9): phase for the positive Weyl convention.
  • idealDisplacementKernel: PDF p.29 (D9): weak delta-comb kernel, evaluated against a test function.
  • smearedWave: PDF pp.18,29 (7.7c),(D10): actual Gaussian-smeared comb wavefunction. The factor sqrt(2/sqrt(pi)) is the D9 weak-kernel normalization, not unit norm.
  • displaceWave: PDF p.29 (D10): D_z on position wavefunctions under the positive Weyl convention.
  • smearedNorm: PDF p.29: exact, generally nonunit, squared norm of the smeared vector.
  • normalizedSmearedWave: PDF pp.18,29: individual normalization; finite-delta logical orthogonality is NOT assumed.
  • smearedKernelIntegral: PDF p.29 (D10): convolution after evaluating the weak delta kernel; the factor 1/(pi*d^2/2) and both ordered Weyl factors are retained.
  • smearedKernelSum: PDF p.29 (D11): exact finite-smearing lattice kernel, including the envelope.
  • hermitePhys: PDF p.18 (7.9): physicists' Hermite polynomial for the oscillator basis.
  • hermiteWave: PDF p.18: normalized position wavefunction of the nth Fock state.
  • gkpFock: PDF p.18 (7.7c): actual Fock coefficients of the individually normalized smeared wave.
  • LatticeShape: PDF pp.20--21: the two lattice geometries explicitly covered by the bound.
  • gkpNormalizerPoint: PDF pp.20--21: lattice normalizer step, covolume pi/2; stabilizers have twice the steps.
  • nearestLattice: PDF p.20 (7.22): syndrome rounding in the normalizer lattice. A fixed classical choice only resolves measure-zero ties; it is independent of the logical input.
  • correctableCell: PDF p.20: central Voronoi cell, including ties.
  • modularSyndrome: PDF p.20 (7.22): measured residual displacement after the chosen ideal correction.
  • logicalCoset: PDF p.20: normalizer modulo stabilizer (2L), i.e. the residual logical Pauli coset.
  • amplifiedLossLaw: PDF p.20 (7.21): actual Gaussian measure, with no probability axiom assumed.
  • cosetProbability: PDF pp.20--21: exact probabilities of the four logical cosets, including identity translates.
  • cosetPauli: PDF p.20: logical X^a Z^b action after ideal modular correction.
  • idealRecoveredChannel: PDF p.20 (7.22): logical channel constructed from syndrome and lattice cosets, with the amplification noise already included. It has no physical ideal-state encoder.
  • centralSuccess: PDF p.20 (7.23): guaranteed central square-cell probability.
  • centralFailure: PDF p.21 (7.24): central-cell failure, which can exceed logical entanglement infidelity.
  • noJump: PDF p.22 (8.3): no-jump damping, with jump-rate scalar factors deliberately omitted.
  • jumpTrajectory: PDF p.22 (8.3): chronological product of actual annihilations and damping intervals.
  • trajectoryScalar: PDF p.22: the scalar relating the trajectory to the integrated loss Kraus operator.
Definition code
import Mathlib
import Definitions.Def_QAlgorithms_Defs_QuditChannelFidelities
import Definitions.Def_QAlgorithms_Defs_Ensemble
import Definitions.Def_QAlgorithms_Defs_Gates
import Definitions.Def_QAlgorithms_Papers_Independent_Advance50_P29
import Definitions.Def_QAlgorithms_Papers_Independent_Frontier18_P14
import Definitions.Def_QAlgorithms_Papers_Independent_Information100_P56
import Definitions.Def_QAlgorithms_Papers_Independent_Dependency50_P40

/-!
Statements from Albert et al., arXiv:1708.05010v3. All page references are PDF pages.
Oscillator formulas use the complete number basis, never a numerical Fock cutoff.
A coordinate sequence is admitted as a Hilbert vector only with `fockDomain`;
unbounded powers carry `annDomain` or `fockMomentDomain`. Ideal combs instead act on
Schwartz test functions. No ideal comb is declared to be a unit Hilbert vector.
-/

noncomputable section

namespace QAlgorithms.Papers.Independent.Intakenext100.P65

open MeasureTheory Filter
open QAlgorithms.Papers.Independent.Frontier18.P14 (displacementEntry)
open QAlgorithms.Papers.Independent.Dependency50.P40 (outputFirstChoi)
open scoped ComplexConjugate Matrix Topology

attribute [local instance] Classical.propDecidable
attribute [local instance] Classical.decEq

/-- PDF pp.2--3: number-basis coordinates, prior to imposing the Hilbert domain. -/
abbrev FockCoordinates := ℕ → ℂ

/-- PDF p.2: the full oscillator Hilbert domain, expressed in its countable ONB. -/
def fockDomain (v : FockCoordinates) : Prop := Summable (fun n => Complex.normSq (v n))

/-- PDF p.2: squared Hilbert norm on the square-summable domain. -/
def fockNormSq (v : FockCoordinates) : ℝ := ∑' n, Complex.normSq (v n)

/-- PDF p.2: inner product, conjugate-linear in the first entry. -/
def fockInner (v w : FockCoordinates) : ℂ := ∑' n, conj (v n) * w n

/-- PDF pp.2,4: a finite photon-number moment has a summability domain. -/
def fockMomentDomain (v : FockCoordinates) (k : ℕ) : Prop :=
  Summable (fun n : ℕ => (n : ℝ) ^ k * Complex.normSq (v n))

/-- PDF pp.4,18: kth number moment, only used on `fockMomentDomain`. -/
def numberMoment (v : FockCoordinates) (k : ℕ) : ℝ :=
  ∑' n : ℕ, (n : ℝ) ^ k * Complex.normSq (v n)

/-- PDF p.4 (2.1): literal half-sum of the two logical number moments. -/
def codeMoment (B : Fin 2 → FockCoordinates) (k : ℕ) : ℝ :=
  (numberMoment (B 0) k + numberMoment (B 1) k) / 2

/-- PDF p.2: coefficient of a^k, including its factorial normalization. -/
def annPower (k : ℕ) (v : FockCoordinates) : FockCoordinates :=
  fun n => (Real.sqrt ((Nat.factorial (n + k) : ℝ) / Nat.factorial n) : ℂ) * v (n + k)

/-- PDF p.2: natural domain of the unbounded annihilation power. -/
def annDomain (k : ℕ) (v : FockCoordinates) : Prop := fockDomain v ∧ fockDomain (annPower k v)

/-- PDF p.2 (1.4): the exact loss coefficient for input occupation n, zero below l. -/
def lossCoefficient (g : ℝ) (l n : ℕ) : ℝ :=
  if l ≤ n then Real.sqrt (Nat.choose n l : ℝ) * Real.sqrt g ^ l *
    Real.sqrt (1 - g) ^ (n - l) else 0

/-- PDF p.2 (1.4): loss Kraus action on the full number basis. -/
def lossVector (g : ℝ) (l : ℕ) (v : FockCoordinates) : FockCoordinates :=
  fun n => (lossCoefficient g l (n + l) : ℂ) * v (n + l)

/-- PDF p.8 (4.1): actual compression of E_l^dagger E_k, not a free QEC matrix. -/
def qecBlock (B : Fin 2 → FockCoordinates) (g : ℝ) (l k : ℕ) : Matrix (Fin 2) (Fin 2) ℂ :=
  fun i j => fockInner (lossVector g l (B i)) (lossVector g k (B j))

/-- PDF p.8 (4.4): coefficient of the identity. -/
def qecC (B : Fin 2 → FockCoordinates) (g : ℝ) (l k : ℕ) : ℝ :=
  ((qecBlock B g l k 0 0 + qecBlock B g l k 1 1) / 2).re

/-- PDF p.8 (4.4): coefficient of logical Z in the chosen basis. -/
def qecZ (B : Fin 2 → FockCoordinates) (g : ℝ) (l k : ℕ) : ℝ :=
  ((qecBlock B g l k 0 0 - qecBlock B g l k 1 1) / 2).re

/-- PDF p.8 (4.4): coefficient of logical X, including nonsymmetric off-diagonal blocks. -/
def qecX (B : Fin 2 → FockCoordinates) (g : ℝ) (l k : ℕ) : ℝ :=
  ((qecBlock B g l k 0 1 + qecBlock B g l k 1 0) / 2).re

/-- PDF p.3 (1.8): I,X,Y,Z, using the accepted Bool-indexed Pauli matrices
with Fin 2 labels 0,1 mapped to false,true. -/
def liouvilleBasis : Fin 4 → Matrix (Fin 2) (Fin 2) ℂ :=
  ![1, (fun i j => pauliX (decide (i.val = 1)) (decide (j.val = 1))),
    (fun i j => pauliY (decide (i.val = 1)) (decide (j.val = 1))),
    (fun i j => pauliZ (decide (i.val = 1)) (decide (j.val = 1)))]

/-- PDF p.3 (1.8): Pauli Liouville matrix; this is distinct from the Choi matrix. -/
def liouville (T : Superop (Fin 2) (Fin 2)) : Matrix (Fin 4) (Fin 4) ℂ :=
  fun k l => ((liouvilleBasis k * T (liouvilleBasis l)).trace) / 2

/-- PDF p.7 (3.3): trace over the reference, leaving the channel's output marginal. -/
def traceReference (R : Matrix (Fin 2 × Fin 2) (Fin 2 × Fin 2) ℂ) : Matrix (Fin 2) (Fin 2) ℂ :=
  fun i j => ∑ r, R (i, r) (j, r)

/-- PDF p.7 (3.3): reverse coherentCoordinates information, without capacity optimization.
The frozen output-first Choi operator is unnormalized; multiplying by 1/2
gives the trace-one qubit Choi state in (1.9). -/
def reverseInformation (T : Superop (Fin 2) (Fin 2)) : ℝ :=
  vnEntropy (traceReference ((1 / 2 : ℂ) • outputFirstChoi T)) - vnEntropy ((1 / 2 : ℂ) • outputFirstChoi T)

/-- PDF p.7 (3.4): squared Frobenius norm from accepted Advance50.P29,
specialized to the effective qubit channel. -/
def qubitFrobeniusSq (A : Matrix (Fin 2) (Fin 2) ℂ) : ℝ :=
  QAlgorithms.Papers.Independent.Advance50.P29.frobeniusSq A

/-- PDF p.24 (A2): minimum over normalized pure inputs; compactness and CPTP
supply a nonempty bounded set, with attainment asserted below. -/
def worstPureFidelity (d : ℕ) (T : Superop (Fin d) (Fin d)) : ℝ :=
  sInf {r : ℝ | ∃ v : EuclideanSpace ℂ (Fin d), ‖v‖ = 1 ∧
    r = effectProb (pureDensity v) (T (pureDensity v))}

/-- PDF p.9 (5.3): periodic residue projector, allowing signed residue labels. -/
def residueProject (S : ℕ) (r : ℤ) (v : FockCoordinates) : FockCoordinates :=
  fun n => if (n : ℤ) % (2 * (S + 1) : ℕ) = r % (2 * (S + 1) : ℕ) then v n else 0

/-- PDF p.9: normalized coherentCoordinates-state coefficients for real amplitude a. -/
def coherentCoordinates (a : ℝ) : FockCoordinates :=
  QAlgorithms.Papers.Independent.Frontier18.P14.coherentCoefficient (a : ℂ)

/-- PDF p.9 (5.2): the actual projected coherentCoordinates-state probability. -/
def catNorm (a : ℝ) (S : ℕ) (r : ℤ) : ℝ :=
  fockNormSq (residueProject S r (coherentCoordinates a))

/-- PDF p.9 (5.1): normalized projected coherentCoordinates coordinates; a>0 in every use. -/
def catCodeCoordinates (a : ℝ) (S : ℕ) (mu : Fin 2) : FockCoordinates :=
  fun n => residueProject S ((S + 1) * mu.val : ℕ) (coherentCoordinates a) n /
    (Real.sqrt (catNorm a S ((S + 1) * mu.val : ℕ)) : ℂ)

/-- PDF p.9 (5.7): bounded number-phase rotation. -/
def residueCheck (S : ℕ) (v : FockCoordinates) : FockCoordinates :=
  fun n => Complex.exp (Complex.I * (2 * Real.pi * (n : ℝ) / (S + 1))) * v n

/-- PDF p.9 following (5.7): the specified square root of the check is logical Z. -/
def residueLogicalZ (S : ℕ) (v : FockCoordinates) : FockCoordinates :=
  fun n => Complex.exp (Complex.I * (Real.pi * (n : ℝ) / (S + 1))) * v n

/-- PDF p.9 (5.5): Fourier projection using exact coherentCoordinates-state overlaps. -/
def catNormFourier (a : ℝ) (S : ℕ) (mu : Fin 2) : ℂ :=
  (1 / (2 * (S + 1) : ℂ)) * ∑ s ∈ Finset.range (2 * (S + 1)),
    (-1 : ℂ) ^ (mu.val * s) *
      Complex.exp ((a ^ 2 : ℝ) * (Complex.exp (Complex.I * (Real.pi * (s : ℝ) / (S + 1))) - 1))

/-- PDF pp.9--11: probability ratios with signed residues, before taking any limit. -/
def catLossRatio (a : ℝ) (S : ℕ) (g : ℝ) (l : ℕ) (mu : Fin 2) : ℝ :=
  catNorm (a * Real.sqrt (1 - g)) S ((((S + 1) * mu.val : ℕ) : ℤ) - (l : ℤ)) /
    catNorm a S ((S + 1) * mu.val : ℕ)

/-- PDF p.10 (5.12): Poisson weight whose mean g*a^2 varies with a. -/
def poissonWeight (a g : ℝ) (l : ℕ) : ℝ :=
  (g * a ^ 2) ^ l * Real.exp (-g * a ^ 2) / Nat.factorial l

/-- PDF p.14 (6.11): finite spin coefficients, indexed by twice the spin. -/
def antipodal (M : ℕ) (mu : Fin 2) (m : Fin (M + 1)) : ℂ :=
  (-1 : ℂ) ^ (mu.val * m.val) * (Real.sqrt (Nat.choose M m.val) / Real.sqrt ((2 : ℝ) ^ M) : ℝ)

/-- PDF p.11 (6.1): direct full-Fock binomialCodeCoordinates coordinates. -/
def binomialCodeCoordinates (N S : ℕ) (mu : Fin 2) : FockCoordinates :=
  fun n => if h : n % (S + 1) = 0 ∧ n / (S + 1) ≤ N + 1 then
    antipodal (N + 1) mu ⟨n / (S + 1), Nat.lt_succ_of_le h.2⟩ else 0

/-- PDF p.12 (6.4): the positive/negative logical combinations. -/
def binomialParity (N S : ℕ) (mu : Fin 2) : FockCoordinates :=
  fun n => (binomialCodeCoordinates N S 0 n + (-1 : ℂ) ^ mu.val * binomialCodeCoordinates N S 1 n) / (Real.sqrt 2 : ℂ)

/-- PDF p.12 (6.5): exact generating polynomial, with plus/minus sign. -/
def binomialGenerator (N S : ℕ) (minus : Bool) : Polynomial ℝ :=
  ((Polynomial.C (1 / 2)) *
    (1 + Polynomial.C (if minus then -1 else 1) * Polynomial.X ^ (S + 1))) ^ (N + 1)

/-- PDF p.14: standard spin-x matrix, on m=0,...,M for spin M/2. -/
def spinX (M : ℕ) : Matrix (Fin (M + 1)) (Fin (M + 1)) ℂ :=
  fun i j => if i.val = j.val + 1 then
      (Real.sqrt (((j.val + 1 : ℕ) : ℝ) * (M - j.val)) / 2 : ℝ)
    else if j.val = i.val + 1 then
      (Real.sqrt (((i.val + 1 : ℕ) : ℝ) * (M - i.val)) / 2 : ℝ) else 0

/-- PDF pp.14--15: transport of a coefficient vector to one mode. -/
def embedOne (M S : ℕ) (v : Fin (M + 1) → ℂ) : FockCoordinates :=
  fun n => ∑ m, if n = (S + 1) * m.val then v m else 0

/-- PDF p.15 Fig.7(a), p.27: two-mode transport with the spacing in BOTH occupations.
Printed (6.16) drops a spacing factor; use the explicit preceding basis map. -/
def embedTwo (M S : ℕ) (v : Fin (M + 1) → ℂ) : ℕ × ℕ → ℂ :=
  fun n => ∑ m, if n = ((S + 1) * (M - m.val), (S + 1) * m.val) then v m else 0

/-- PDF p.15: accepted Information100.P22.dicke in Fin 2 word coordinates;
Bool false,true correspond to 0,1. Infeasible excitation weights give zero. -/
def qubitDickeCoordinates (K r : ℕ) (x : Fin K → Fin 2) : ℂ :=
  QAlgorithms.Papers.Independent.Information100.P22.dicke (I := Fin K) r
    (fun i => decide ((x i).val = 1))

/-- PDF p.15 (6.14): the spaced Dicke transport (K >= M(S+1) at use sites). -/
def embedDicke (M S K : ℕ) (v : Fin (M + 1) → ℂ) (x : Fin K → Fin 2) : ℂ :=
  ∑ m, v m * qubitDickeCoordinates K ((S + 1) * m.val) x

/-- PDF p.17: even/odd spin generator on the stated seven-dimensional invariant space. -/
def smallGenerator : Matrix (Fin 7) (Fin 7) ℂ :=
  let even : Matrix (Fin 7) (Fin 4) ℂ := fun n m => if n.val = 2 * m.val then 1 else 0
  let odd : Matrix (Fin 7) (Fin 3) ℂ := fun n m => if n.val = 2 * m.val + 1 then 1 else 0
  8 • (even * spinX 3 * evenᴴ) ^ 2 + (odd * spinX 2 * oddᴴ) ^ 2

/-- PDF p.17 (6.21): a=8,b=5, with identity extension outside the invariant space. -/
def smallUnitary : Matrix (Fin 7) (Fin 7) ℂ :=
  NormedSpace.exp (((((2 * Real.pi / 5 : ℝ) : ℂ) * Complex.I) • smallGenerator))

/-- PDF p.17 Tables II--III: actual normalized code, dephasing residual, loss, and failure vectors.
Residual normalization is sqrt(3), correcting the false coefficient 2 in (6.20). -/
def smallBasis (j : Fin 7) (n : Fin 7) : ℂ :=
  if h0 : j.val < 2 then binomialCodeCoordinates 2 1 ⟨j.val, by omega⟩ n.val
  else if h1 : j.val < 4 then
    (((n.val : ℝ) - 3) / Real.sqrt 3 : ℝ) * binomialCodeCoordinates 2 1 ⟨j.val - 2, by omega⟩ n.val
  else if h2 : j.val < 6 then
    annPower 1 (binomialCodeCoordinates 2 1 ⟨j.val - 4, by omega⟩) n.val / (Real.sqrt 3 : ℂ)
  else if n.val = 1 then (1 / Real.sqrt 2 : ℝ)
  else if n.val = 5 then (-1 / Real.sqrt 2 : ℝ) else 0

/-- Small exact sanity cases for the new coordinate constructors. -/

example : lossCoefficient (0 : ℝ) 1 0 = 0 := by simp [lossCoefficient]

example (v : FockCoordinates) : residueProject 0 0 v 0 = v 0 := by simp [residueProject]

example : coherentCoordinates 0 0 = 1 := by simp [coherentCoordinates,
  QAlgorithms.Papers.Independent.Frontier18.P14.coherentCoefficient]

example : binomialCodeCoordinates 0 0 0 0 = (1 / Real.sqrt 2 : ℝ) := by
  simp [binomialCodeCoordinates, antipodal]

example : (spinX 0) 0 0 = 0 := by simp [spinX]

example (v : Fin 1 → ℂ) : embedTwo 0 0 v (0, 0) = v 0 := by simp [embedTwo]

/-- PDF p.28 (C4): totalQuditExcitation of an M-qudit computational word. -/
def totalQuditExcitation {d M : ℕ} (x : Fin M → Fin d) : ℕ := ∑ i, (x i).val

/-- PDF p.27 (C2),(C4): number of words of a given total totalQuditExcitation. -/
def extendedCount (d M m : ℕ) : ℕ :=
  (Finset.univ.filter (fun x : Fin M → Fin d => totalQuditExcitation x = m)).card

/-- PDF p.27 (C4): normalized extended binomialCodeCoordinates vector, distinct from a general qudit Dicke vector. -/
def extendedVector (d M m : ℕ) (x : Fin M → Fin d) : ℂ :=
  if totalQuditExcitation x = m then (1 / Real.sqrt (extendedCount d M m) : ℝ) else 0

/-- PDF p.28 (C8): positive Fourier phase, matching the explicit backwards shift (C7). -/
def quditFourierFactor (d : ℕ) (mu m : ℕ) : ℂ :=
  Complex.exp (Complex.I * (2 * Real.pi * (mu : ℝ) * (m : ℝ) / d))

/-- PDF p.27 (C3): totalQuditExcitation-basis coefficients, including the endpoint (d-1)M. -/
def extendedCoefficients (d M : ℕ) (mu : Fin d) (m : Fin ((d - 1) * M + 1)) : ℂ :=
  quditFourierFactor d mu.val m.val *
    (Real.sqrt (extendedCount d M m.val) / Real.sqrt ((d : ℝ) ^ M) : ℝ)

/-- PDF p.27 (C3): the new multi-qudit code, in extended-binomialCodeCoordinates coordinates. -/
def permCode (d M : ℕ) (mu : Fin d) (x : Fin M → Fin d) : ℂ :=
  ∑ m : Fin ((d - 1) * M + 1), extendedCoefficients d M mu m * extendedVector d M m.val x

/-- PDF p.28 (C9): tensor product of M identical Fourier vectors. -/
def fourierProduct (d M : ℕ) (mu : Fin d) (x : Fin M → Fin d) : ℂ :=
  ∏ i, quditFourierFactor d mu.val (x i).val / (Real.sqrt d : ℂ)

/-- PDF p.28 (C7): the input label one step above the output label. -/
def nextLabel {d : ℕ} (hd : 0 < d) (j : Fin d) : Fin d :=
  ⟨(j.val + 1) % d, Nat.mod_lt _ hd⟩

/-- PDF p.28 (C10): average of the backwards cyclic shifts on individual qudits. -/
def collectiveShift (d M : ℕ) (hd : 0 < d) :
    Matrix (Fin M → Fin d) (Fin M → Fin d) ℂ :=
  fun x y => (1 / (M : ℂ)) * ∑ k : Fin M,
    if y = Function.update x k (nextLabel hd (x k)) then 1 else 0

/-- PDF p.28: compression to the extended-binomialCodeCoordinates span, of dimension (d-1)M+1. -/
def compressedShift (d M : ℕ) (hd : 0 < d) :
    Matrix (Fin ((d - 1) * M + 1)) (Fin ((d - 1) * M + 1)) ℂ :=
  let V : Matrix (Fin M → Fin d) (Fin ((d - 1) * M + 1)) ℂ :=
    fun x m => extendedVector d M m.val x
  Vᴴ * collectiveShift d M hd * V

/-- PDF p.28 (D1b): the positive Weyl convention. -/
def weylPhase (a b : ℂ) : ℂ := Complex.exp ((a * conj b - conj a * b) / 2)

/-- PDF p.2 (1.4): number-basis kernel of a single loss Kraus operator. -/
def lossEntry (g : ℝ) (l n m : ℕ) : ℂ :=
  if m = n + l then (lossCoefficient g l m : ℂ) else 0

/-- PDF p.19 (7.12): genuine product kernel, before the displacement expansion. -/
def lossProduct (g : ℝ) (l k n m : ℕ) : ℂ :=
  ∑' j, conj (lossEntry g l j n) * lossEntry g k j m

/-- PDF p.19 (7.12): integrand INCLUDING the factor 1/pi, on complex-plane Lebesgue measure. -/
def lossDisplacementIntegrand (g : ℝ) (l k n m : ℕ) (z : ℂ) : ℂ :=
  (Real.exp (-(1 - g) * Complex.normSq z / 2) / Real.pi : ℝ) *
    displacementEntry (conj z) l k * displacementEntry ((Real.sqrt g : ℂ) * z) n m

section FockDensityOrder

open scoped ComplexOrder

/-- PDF pp.2,20: positive trace-one full Fock kernels. Positivity of every finite
principal compression and summability of the nonnegative diagonal characterize
positive trace-class operators; there is no fixed cutoff. -/
def fockDensity (R : ℕ → ℕ → ℂ) : Prop :=
  (∀ d : ℕ, Matrix.PosSemidef ((fun i j : Fin d => R i.val j.val) : Matrix (Fin d) (Fin d) ℂ)) ∧
    Summable (fun n => (R n n).re) ∧ (∑' n, (R n n).re) = 1

end FockDensityOrder

/-- PDF p.2: pure loss, by the countable Kraus sum on actual Fock kernels. -/
def lossChannel (g : ℝ) (R : ℕ → ℕ → ℂ) (n m : ℕ) : ℂ :=
  ∑' l, (lossCoefficient g l (n + l) * lossCoefficient g l (m + l) : ℝ) * R (n + l) (m + l)

/-- PDF p.20 (7.20): amplification after loss; Kraus operators sqrt(1-g) E_l^dagger. -/
def amplifierChannel (g : ℝ) (R : ℕ → ℕ → ℂ) (n m : ℕ) : ℂ :=
  ∑ l ∈ Finset.range (min n m + 1),
    ((1 - g) * lossCoefficient g l n * lossCoefficient g l m : ℝ) * R (n - l) (m - l)

/-- PDF p.20 (7.21): displacement conjugation, evaluated on all matrix entries. -/
def displaceDensity (z : ℂ) (R : ℕ → ℕ → ℂ) (n m : ℕ) : ℂ :=
  ∑' j, ∑' k, displacementEntry z n j * R j k * conj (displacementEntry z m k)

/-- PDF p.20 (7.21): density of the amplified-loss displacement distribution on C. -/
def amplifiedLossDensity (g : ℝ) (z : ℂ) : ℝ :=
  (1 - g) / (Real.pi * g) * Real.exp (-(1 - g) / g * Complex.normSq z)

/-- PDF p.18 (7.5): distributional square-code comb acting linearly on tests. -/
def idealComb (mu : Fin 2) (f : ℝ → ℂ) : ℂ :=
  ∑' n : ℤ, f (Real.sqrt Real.pi * (2 * (n : ℝ) + mu.val))

/-- PDF pp.18,29 (7.6),(D5): Gaussian-Abel regularized coherentCoordinates-lattice sum.
The ordered product is D_real D_imag; it is evaluated on the vacuum in position coordinates. -/
def regularizedLattice (mu : Fin 2) (r x : ℝ) : ℂ :=
  ∑' n : ℤ × ℤ,
    let q := Real.sqrt Real.pi * (2 * (n.1 : ℝ) + mu.val)
    (Real.exp (-r * ((n.1 : ℝ) ^ 2 + (n.2 : ℝ) ^ 2)) *
      Real.rpow Real.pi (-1 / 4 : ℝ) * Real.exp (-(x - q) ^ 2 / 2) : ℝ) *
        Complex.exp (Complex.I * (Real.sqrt Real.pi * (n.2 : ℝ) * (x - q)))

/-- PDF p.29 (D9): normalizer-lattice displacement, with signed mu-nu offset. -/
def kernelLattice (mu nu : Fin 2) (n : ℤ × ℤ) : ℂ :=
  (Real.sqrt (Real.pi / 2) : ℂ) *
    (((2 * (n.1 : ℝ) + (mu.val : ℝ) - nu.val) : ℂ) + Complex.I * (n.2 : ℝ))

/-- PDF p.29 (D9): phase for the positive Weyl convention. -/
def kernelPhase (mu nu : Fin 2) (n : ℤ × ℤ) : ℂ :=
  Complex.exp (Complex.I * (Real.pi * ((n.1 : ℝ) + ((mu.val : ℝ) + nu.val) / 2) * (n.2 : ℝ)))

/-- PDF p.29 (D9): weak delta-comb kernel, evaluated against a test function. -/
def idealDisplacementKernel (mu nu : Fin 2) (f : ℂ → ℂ) : ℂ :=
  ∑' n : ℤ × ℤ, kernelPhase mu nu n * f (kernelLattice mu nu n)

/-- PDF pp.18,29 (7.7c),(D10): actual Gaussian-smeared comb wavefunction.
The factor sqrt(2/sqrt(pi)) is the D9 weak-kernel normalization, not unit norm. -/
def smearedWave (mu : Fin 2) (d x : ℝ) : ℂ :=
  (Real.sqrt (2 / Real.sqrt Real.pi) : ℂ) * ∑' n : ℤ,
    let q := Real.sqrt Real.pi * (2 * (n : ℝ) + mu.val)
    (Real.exp (-(x - q) ^ 2 / (2 * d ^ 2) - d ^ 2 * (x + q) ^ 2 / 8) : ℝ)

/-- PDF p.29 (D10): D_z on position wavefunctions under the positive Weyl convention. -/
def displaceWave (z : ℂ) (v : ℝ → ℂ) (x : ℝ) : ℂ :=
  Complex.exp (Complex.I * ((Real.sqrt 2 * z.im * x - z.re * z.im) : ℝ)) *
    v (x - Real.sqrt 2 * z.re)

/-- PDF p.29: exact, generally nonunit, squared norm of the smeared vector. -/
def smearedNorm (mu : Fin 2) (d : ℝ) : ℝ := ∫ x : ℝ, Complex.normSq (smearedWave mu d x)

/-- PDF pp.18,29: individual normalization; finite-delta logical orthogonality is NOT assumed. -/
def normalizedSmearedWave (mu : Fin 2) (d x : ℝ) : ℂ :=
  smearedWave mu d x / (Real.sqrt (smearedNorm mu d) : ℂ)

/-- PDF p.29 (D10): convolution after evaluating the weak delta kernel; the factor
1/(pi*d^2/2) and both ordered Weyl factors are retained. -/
def smearedKernelIntegral (mu nu : Fin 2) (d : ℝ) (z : ℂ) : ℂ :=
  ∑' n : ℤ × ℤ, kernelPhase mu nu n *
    ∫ b : ℂ,
      (Real.exp (-(Complex.normSq b + Complex.normSq (kernelLattice mu nu n + b - z)) / d ^ 2) /
        (Real.pi * d ^ 2 / 2) : ℝ) *
        weylPhase (-b) z * weylPhase (z - b) (kernelLattice mu nu n + b - z)

/-- PDF p.29 (D11): exact finite-smearing lattice kernel, including the envelope. -/
def smearedKernelSum (mu nu : Fin 2) (d : ℝ) (z : ℂ) : ℂ :=
  ∑' n : ℤ × ℤ, kernelPhase mu nu n *
    (Real.exp (-Complex.normSq (z - kernelLattice mu nu n) / (2 * d ^ 2) -
      d ^ 2 * Complex.normSq (z + kernelLattice mu nu n) / 8) : ℝ)

/-- PDF p.18 (7.9): physicists' Hermite polynomial for the oscillator basis. -/
def hermitePhys : ℕ → Polynomial ℝ
  | 0 => 1
  | n + 1 => 2 * Polynomial.X * hermitePhys n - Polynomial.derivative (hermitePhys n)

/-- PDF p.18: normalized position wavefunction of the nth Fock state. -/
def hermiteWave (n : ℕ) (x : ℝ) : ℝ :=
  (hermitePhys n).eval x * Real.exp (-x ^ 2 / 2) /
    Real.sqrt ((2 : ℝ) ^ n * Nat.factorial n * Real.sqrt Real.pi)

/-- PDF p.18 (7.7c): actual Fock coefficients of the individually normalized smeared wave. -/
def gkpFock (d : ℝ) (mu : Fin 2) : FockCoordinates :=
  fun n => ∫ x : ℝ, (hermiteWave n x : ℂ) * normalizedSmearedWave mu d x

/-- PDF pp.20--21: the two lattice geometries explicitly covered by the bound. -/
inductive LatticeShape where
  | square
  | hexagonal
  deriving DecidableEq

/-- PDF pp.20--21: lattice normalizer step, covolume pi/2; stabilizers have twice the steps. -/
def gkpNormalizerPoint (shape : LatticeShape) (n : ℤ × ℤ) : ℂ :=
  match shape with
  | .square => (Real.sqrt (Real.pi / 2) : ℂ) * ((n.1 : ℂ) + Complex.I * (n.2 : ℂ))
  | .hexagonal => (Real.sqrt (Real.pi / Real.sqrt 3) : ℂ) *
      ((n.1 : ℂ) + (n.2 : ℂ) * (1 / 2 + Complex.I * (Real.sqrt 3 / 2 : ℝ)))

/-- PDF p.20 (7.22): syndrome rounding in the normalizer lattice.
A fixed classical choice only resolves measure-zero ties; it is independent of the logical input. -/
def nearestLattice (shape : LatticeShape) (z : ℂ) : ℤ × ℤ :=
  Classical.epsilon (fun n => ∀ m : ℤ × ℤ, ‖z - gkpNormalizerPoint shape n‖ ≤ ‖z - gkpNormalizerPoint shape m‖)

/-- PDF p.20: central Voronoi cell, including ties. -/
def correctableCell (shape : LatticeShape) : Set ℂ :=
  {z | ∀ n : ℤ × ℤ, ‖z‖ ≤ ‖z - gkpNormalizerPoint shape n‖}

/-- PDF p.20 (7.22): measured residual displacement after the chosen ideal correction. -/
def modularSyndrome (shape : LatticeShape) (z : ℂ) : ℂ := z - gkpNormalizerPoint shape (nearestLattice shape z)

/-- PDF p.20: normalizer modulo stabilizer (2L), i.e. the residual logical Pauli coset. -/
def logicalCoset (shape : LatticeShape) (z : ℂ) : Fin 2 × Fin 2 :=
  (⟨((nearestLattice shape z).1 % 2).toNat, by omega⟩,
   ⟨((nearestLattice shape z).2 % 2).toNat, by omega⟩)

/-- PDF p.20 (7.21): actual Gaussian measure, with no probability axiom assumed. -/
def amplifiedLossLaw (g : ℝ) : Measure ℂ :=
  volume.withDensity (fun z => ENNReal.ofReal (amplifiedLossDensity g z))

/-- PDF pp.20--21: exact probabilities of the four logical cosets, including identity translates. -/
def cosetProbability (shape : LatticeShape) (g : ℝ) (c : Fin 2 × Fin 2) : ℝ :=
  (amplifiedLossLaw g {z | logicalCoset shape z = c}).toReal

/-- PDF p.20: logical X^a Z^b action after ideal modular correction. -/
def cosetPauli (c : Fin 2 × Fin 2) : Matrix (Fin 2) (Fin 2) ℂ :=
  (liouvilleBasis 1) ^ c.1.val * (liouvilleBasis 3) ^ c.2.val

/-- PDF p.20 (7.22): logical channel constructed from syndrome and lattice cosets,
with the amplification noise already included. It has no physical ideal-state encoder. -/
def idealRecoveredChannel (shape : LatticeShape) (g : ℝ) : Superop (Fin 2) (Fin 2) :=
  krausApply (fun c : Fin 2 × Fin 2 => (Real.sqrt (cosetProbability shape g c) : ℂ) • cosetPauli c)

/-- PDF p.20 (7.23): guaranteed central square-cell probability. -/
def centralSuccess (g : ℝ) : ℝ :=
  (amplifiedLossLaw g {z | |z.re| ≤ Real.sqrt (Real.pi / 8) ∧ |z.im| ≤ Real.sqrt (Real.pi / 8)}).toReal

/-- PDF p.21 (7.24): central-cell failure, which can exceed logical entanglement infidelity. -/
def centralFailure (g : ℝ) : ℝ :=
  (amplifiedLossLaw g {z | ¬ (|z.re| ≤ Real.sqrt (Real.pi / 8) ∧ |z.im| ≤ Real.sqrt (Real.pi / 8))}).toReal

/-- PDF p.22 (8.3): no-jump damping, with jump-rate scalar factors deliberately omitted. -/
def noJump (k t : ℝ) (v : FockCoordinates) : FockCoordinates :=
  fun n => (Real.exp (-k * t * n / 2) : ℂ) * v n

/-- PDF p.22 (8.3): chronological product of actual annihilations and damping intervals. -/
def jumpTrajectory (k t : ℝ) {l : ℕ} (times : Fin l → ℝ) (v : FockCoordinates) : FockCoordinates :=
  let final := (List.ofFn times).foldl
    (fun (s : ℝ × FockCoordinates) tau => (tau, annPower 1 (noJump k (tau - s.1) s.2))) (0, v)
  noJump k (t - final.1) final.2

/-- PDF p.22: the scalar relating the trajectory to the integrated loss Kraus operator. -/
def trajectoryScalar (k : ℝ) {l : ℕ} (times : Fin l → ℝ) (g : ℝ) : ℝ :=
  Real.sqrt (Nat.factorial l) * Real.exp (-k * (∑ i, times i) / 2) / Real.sqrt g ^ l

/-- Sanity tests: endpoints, orientations, and empty trajectories. -/

example (d : ℕ) (mu m : ℕ) : quditFourierFactor d mu 0 = 1 := by simp [quditFourierFactor]

example : displacementEntry 0 0 0 = 1 := by simp [QAlgorithms.Papers.Independent.Frontier18.P14.displacementEntry,
  Complex.normSq_zero]

example (mu : Fin 2) : idealComb mu (fun _ => 0) = 0 := by simp [idealComb]

example : kernelLattice 0 0 (0, 0) = 0 := by simp [kernelLattice]

example (g : ℝ) : amplifiedLossDensity g 0 = (1 - g) / (Real.pi * g) := by simp [amplifiedLossDensity]

example (k t : ℝ) (v : FockCoordinates) :
    jumpTrajectory k t (fun i : Fin 0 => Fin.elim0 i) v = noJump k t v := by
  simp [jumpTrajectory]

end QAlgorithms.Papers.Independent.Intakenext100.P65
Source
Performance and structure of single-mode bosonic codes, arXiv:1708.05010v3, https://arxiv.org/abs/1708.05010v3
Read-back

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

QAlgorithms.Papers.Independent.Intakenext100.P65.FockCoordinates

A Fock coordinate sequence is any function v:N→Cv:\mathbb N\to\mathbb Cv:N→C, assigning a complex number vnv_nvn​ to each nonnegative integer nnn. No summability or normalization condition is imposed.

QAlgorithms.Papers.Independent.Intakenext100.P65.LatticeShape

Define a type with exactly two distinct possible values, called square\mathrm{square}square and hexagonal\mathrm{hexagonal}hexagonal, and with decidable equality between its values.

QAlgorithms.Papers.Independent.Intakenext100.P65.amplifiedLossDensity

For every real ggg and complex zzz, define the real number

wg(z)=1−gπgexp⁡ ⁣(−1−gg∣z∣2).w_g(z)=\frac{1-g}{\pi g} \exp\!\left(-\frac{1-g}{g}|z|^2\right).wg​(z)=πg1−g​exp(−g1−g​∣z∣2).

There is no restriction on ggg and no assertion here that this function is a probability density. Division is total, with division by zero equal to zero; in particular, w0(z)=w1(z)=0w_0(z)=w_1(z)=0w0​(z)=w1​(z)=0 for every zzz.

QAlgorithms.Papers.Independent.Intakenext100.P65.amplifiedLossLaw

For every real ggg, define

dg(z)=1−gπgexp⁡ ⁣(−1−gg∣z∣2),z∈C.d_g(z)=\frac{1-g}{\pi g}\exp\!\left(-\frac{1-g}{g}|z|^2\right),\qquad z\in\mathbb C.dg​(z)=πg1−g​exp(−g1−g​∣z∣2),z∈C.

The defined measure is complex-plane volume weighted by the extended nonnegative density max⁡(dg(z),0)\max(d_g(z),0)max(dg​(z),0); equivalently, its value on a measurable set AAA is ∫Amax⁡(dg(z),0) dz\int_A\max(d_g(z),0)\,dz∫A​max(dg​(z),0)dz, interpreted as a nonnegative extended integral. There is no assumption on ggg and no normalization assertion in this definition. Division by zero is interpreted as zero; in particular, the resulting measure is zero whenever g∉(0,1)g\notin(0,1)g∈/(0,1).

QAlgorithms.Papers.Independent.Intakenext100.P65.amplifierChannel

For any real ggg, any complex array R:N×N→CR:\mathbb N\times\mathbb N\to\mathbb CR:N×N→C, and any n,m∈Nn,m\in\mathbb Nn,m∈N, define

(AgR)nm=∑l=0min⁡(n,m)(1−g)cg(l,n)cg(l,m)Rn−l,m−l,cg(l,s)={(sl) (g)l(1−g)s−l,l≤s,0,l>s.(\mathcal A_gR)_{nm} =\sum_{l=0}^{\min(n,m)} (1-g)c_g(l,n)c_g(l,m)R_{n-l,m-l}, \qquad c_g(l,s)= \begin{cases} \sqrt{\binom{s}{l}}\,(\sqrt g)^l(\sqrt{1-g})^{s-l},&l\le s,\\ 0,&l>s. \end{cases}(Ag​R)nm​=l=0∑min(n,m)​(1−g)cg​(l,n)cg​(l,m)Rn−l,m−l​,cg​(l,s)={(ls​)​(g​)l(1−g​)s−l,0,​l≤s,l>s.​

The real coefficients are regarded as complex scalars. This is a finite sum, including when n=0n=0n=0 or m=0m=0m=0, and no restriction on ggg or density assumption on RRR is imposed. All square roots are real square roots, equal to zero on negative inputs; all natural-number subtractions in this sum occur with l≤n,ml\le n,ml≤n,m.

QAlgorithms.Papers.Independent.Intakenext100.P65.annDomain

For every nonnegative integer kkk and every complex sequence v:N→Cv:\mathbb N\to\mathbb Cv:N→C, this predicate requires both of the following real series to be summable:

∑n=0∞∣vn∣2,∑n=0∞∣(n+k)!n! vn+k∣2.\sum_{n=0}^{\infty}|v_n|^2, \qquad \sum_{n=0}^{\infty} \left|\sqrt{\frac{(n+k)!}{n!}}\,v_{n+k}\right|^2.n=0∑∞​∣vn​∣2,n=0∑∞​​n!(n+k)!​​vn+k​​2.

No normalization is required. When k=0k=0k=0, the two requirements coincide; the zero sequence satisfies the predicate for every kkk.

QAlgorithms.Papers.Independent.Intakenext100.P65.annPower

For every nonnegative integer kkk and every complex sequence v:N→Cv:\mathbb N\to\mathbb Cv:N→C, this defines the complex sequence whose nnnth coordinate, for every n≥0n\ge0n≥0, is

(n+k)!n! vn+k,\sqrt{\frac{(n+k)!}{n!}}\,v_{n+k},n!(n+k)!​​vn+k​,

where the factorials are converted to real numbers before division, and the real square root is viewed as a complex scalar. No summability condition is imposed on either sequence. The denominator is always positive, and k=0k=0k=0 gives the original sequence vvv.

QAlgorithms.Papers.Independent.Intakenext100.P65.antipodal

For every natural number MMM, label μ∈{0,1}\mu\in\{0,1\}μ∈{0,1}, and integer index mmm with 0≤m≤M0\le m\le M0≤m≤M, this defines the complex number

(−1)μm(Mm)2M,(-1)^{\mu m}\frac{\sqrt{\binom M m}}{\sqrt{2^M}},(−1)μm2M​(mM​)​​,

with the real square-root ratio regarded as complex. The definition includes M=0M=0M=0, when the only index is m=0m=0m=0 and the value is 111 for either label.

QAlgorithms.Papers.Independent.Intakenext100.P65.binomialCodeCoordinates

For every pair of natural numbers N,SN,SN,S and label μ∈{0,1}\mu\in\{0,1\}μ∈{0,1}, this defines a complex-valued sequence bμ:N→Cb_\mu:\mathbb N\to\mathbb Cbμ​:N→C by

bμ(n)={(−1)μm(N+1m)2N+1,n=(S+1)m for some 0≤m≤N+1,0,otherwise.b_\mu(n)= \begin{cases} (-1)^{\mu m}\dfrac{\sqrt{\binom{N+1}{m}}}{\sqrt{2^{N+1}}}, &n=(S+1)m\text{ for some }0\le m\le N+1,\\ 0,&\text{otherwise}. \end{cases}bμ​(n)=⎩⎨⎧​(−1)μm2N+1​(mN+1​)​​,0,​n=(S+1)m for some 0≤m≤N+1,otherwise.​

Here m=n/(S+1)m=n/(S+1)m=n/(S+1) is the natural-number quotient when the first case applies. Both N=0N=0N=0 and S=0S=0S=0 are included; S+1S+1S+1 is always positive.

QAlgorithms.Papers.Independent.Intakenext100.P65.binomialGenerator

For every pair of natural numbers N,SN,SN,S and Boolean parameter minus\mathrm{minus}minus, this defines the real polynomial

(1+εXS+12)N+1,ε={−1,minus=true,1,minus=false.\left(\frac{1+\varepsilon X^{S+1}}2\right)^{N+1}, \qquad \varepsilon= \begin{cases} -1,&\mathrm{minus}=\mathrm{true},\\ 1,&\mathrm{minus}=\mathrm{false}. \end{cases}(21+εXS+1​)N+1,ε={−1,1,​minus=true,minus=false.​

The definition includes N=0N=0N=0 and S=0S=0S=0.

QAlgorithms.Papers.Independent.Intakenext100.P65.binomialParity

For every pair of natural numbers N,SN,SN,S and label μ∈{0,1}\mu\in\{0,1\}μ∈{0,1}, this defines the complex-valued sequence

pμ(n)=b0(n)+(−1)μb1(n)2,p_\mu(n)=\frac{b_0(n)+(-1)^\mu b_1(n)}{\sqrt2},pμ​(n)=2​b0​(n)+(−1)μb1​(n)​,

where, for ν∈{0,1}\nu\in\{0,1\}ν∈{0,1}, bν((S+1)m)=(−1)νm(N+1m)/2N+1b_\nu((S+1)m)=(-1)^{\nu m}\sqrt{\binom{N+1}{m}/2^{N+1}}bν​((S+1)m)=(−1)νm(mN+1​)/2N+1​ for 0≤m≤N+10\le m\le N+10≤m≤N+1, and bν(n)=0b_\nu(n)=0bν​(n)=0 elsewhere. Equivalently,

pμ((S+1)m)=1+(−1)μ+m2(N+1m)2N+1,p_\mu((S+1)m)= \frac{1+(-1)^{\mu+m}}{\sqrt2} \sqrt{\frac{\binom{N+1}{m}}{2^{N+1}}},pμ​((S+1)m)=2​1+(−1)μ+m​2N+1(mN+1​)​​,

and all other coordinates vanish. The definition includes N=0N=0N=0 and S=0S=0S=0.

QAlgorithms.Papers.Independent.Intakenext100.P65.catCodeCoordinates

For every real aaa, natural number SSS, and label μ∈{0,1}\mu\in\{0,1\}μ∈{0,1}, let

Cμ=∑n=0∞∣{e−a2/2an/n!,n≡(S+1)μ(mod2(S+1)),0,otherwise.∣2.C_\mu=\sum_{n=0}^{\infty} \left| \begin{cases} e^{-a^2/2}a^n/\sqrt{n!},&n\equiv(S+1)\mu\pmod{2(S+1)},\\ 0,&\text{otherwise}. \end{cases} \right|^2.Cμ​=n=0∑∞​​{e−a2/2an/n!​,0,​n≡(S+1)μ(mod2(S+1)),otherwise.​​2.

This defines the complex sequence

(catCodeCoordinates⁡(a,S,μ))n={e−a2/2an/n!,n≡(S+1)μ(mod2(S+1)),0,otherwiseCμ.(\operatorname{catCodeCoordinates}(a,S,\mu))_n =\frac{ \begin{cases} e^{-a^2/2}a^n/\sqrt{n!},&n\equiv(S+1)\mu\pmod{2(S+1)},\\ 0,&\text{otherwise} \end{cases}} {\sqrt{C_\mu}}.(catCodeCoordinates(a,S,μ))n​=Cμ​​{e−a2/2an/n!​,0,​n≡(S+1)μ(mod2(S+1)),otherwise​​.

There is no hypothesis that the denominator is nonzero: complex division by zero is defined to give zero, so a zero denominator makes the entire sequence zero. The definition includes a=0a=0a=0 and S=0S=0S=0, and by itself asserts no normalization property.

QAlgorithms.Papers.Independent.Intakenext100.P65.catLossRatio

For every real a,ga,ga,g, natural numbers S,lS,lS,l, and label μ∈{0,1}\mu\in\{0,1\}μ∈{0,1}, define

C(b,r)=∑n=0∞∣{e−b2/2bn/n!,n≡r(mod2(S+1)),0,otherwise.∣2.C(b,r)=\sum_{n=0}^{\infty} \left| \begin{cases} e^{-b^2/2}b^n/\sqrt{n!},&n\equiv r\pmod{2(S+1)},\\ 0,&\text{otherwise}. \end{cases} \right|^2.C(b,r)=n=0∑∞​​{e−b2/2bn/n!​,0,​n≡r(mod2(S+1)),otherwise.​​2.

The declaration defines

catLossRatio⁡(a,S,g,l,μ)=C ⁣(a1−g, (S+1)μ−l)C ⁣(a, (S+1)μ).\operatorname{catLossRatio}(a,S,g,l,\mu) =\frac{C\!\left(a\sqrt{1-g},\,(S+1)\mu-l\right)} {C\!\left(a,\,(S+1)\mu\right)}.catLossRatio(a,S,g,l,μ)=C(a,(S+1)μ)C(a1−g​,(S+1)μ−l)​.

The residue (S+1)μ−l(S+1)\mu-l(S+1)μ−l is an integer difference and may be negative; it is not truncated at zero. There is no restriction on ggg: the real square root is zero when 1−g<01-g<01−g<0. There is also no nonzero-denominator assumption, and real division by zero gives zero.

QAlgorithms.Papers.Independent.Intakenext100.P65.catNorm

For every real aaa, natural number SSS, and integer rrr, this defines the real number

catNorm⁡(a,S,r)=∑n=0∞∣{e−a2/2an/n!,n≡r(mod2(S+1)),0,otherwise∣2.\operatorname{catNorm}(a,S,r) =\sum_{n=0}^{\infty} \left| \begin{cases} e^{-a^2/2}a^n/\sqrt{n!},&n\equiv r\pmod{2(S+1)},\\ 0,&\text{otherwise} \end{cases} \right|^2.catNorm(a,S,r)=n=0∑∞​​{e−a2/2an/n!​,0,​n≡r(mod2(S+1)),otherwise​​2.

Thus the definition sums the squared absolute values of the coherent-coordinate sequence after retaining only the indicated residue class. No positivity assumption on aaa is required. In particular, at a=0a=0a=0 the value is 111 when r≡0(mod2(S+1))r\equiv0\pmod{2(S+1)}r≡0(mod2(S+1)), and 000 otherwise.

QAlgorithms.Papers.Independent.Intakenext100.P65.catNormFourier

For every real aaa, natural number SSS, and label μ∈{0,1}\mu\in\{0,1\}μ∈{0,1}, this defines the complex number

catNormFourier⁡(a,S,μ)=12(S+1)∑s=02(S+1)−1(−1)μsexp⁡ ⁣[a2(exp⁡ ⁣(i πsS+1)−1)].\operatorname{catNormFourier}(a,S,\mu) =\frac{1}{2(S+1)} \sum_{s=0}^{2(S+1)-1} (-1)^{\mu s} \exp\!\left[ a^2\left( \exp\!\left(\mathrm i\,\frac{\pi s}{S+1}\right)-1 \right)\right].catNormFourier(a,S,μ)=2(S+1)1​s=0∑2(S+1)−1​(−1)μsexp[a2(exp(iS+1πs​)−1)].

All real quantities are embedded in C\mathbb CC where needed. The parameters include a=0a=0a=0 and S=0S=0S=0; the finite sum always has at least two terms. This definition alone asserts neither that its value is real nor that it equals any squared norm.

QAlgorithms.Papers.Independent.Intakenext100.P65.centralFailure

For every real ggg, let μg\mu_gμg​ be complex-plane volume weighted by the extended nonnegative density max⁡{1−gπgexp⁡(−1−gg∣z∣2),0}\max\{\frac{1-g}{\pi g}\exp(-\frac{1-g}{g}|z|^2),0\}max{πg1−g​exp(−g1−g​∣z∣2),0}, with division by zero interpreted as zero. The defined real number is the real conversion of the measure of the complement of the closed square {z:∣Re⁡z∣≤π/8, ∣Im⁡z∣≤π/8}\{z:|\operatorname{Re}z|\le\sqrt{\pi/8},\ |\operatorname{Im}z|\le\sqrt{\pi/8}\}{z:∣Rez∣≤π/8​, ∣Imz∣≤π/8​}; that complement consists of points satisfying ∣Re⁡z∣>π/8|\operatorname{Re}z|>\sqrt{\pi/8}∣Rez∣>π/8​ or ∣Im⁡z∣>π/8|\operatorname{Im}z|>\sqrt{\pi/8}∣Imz∣>π/8​. An infinite measure converts to zero. The definition makes no restriction on ggg and does not itself identify this value with one minus the square’s measure.

QAlgorithms.Papers.Independent.Intakenext100.P65.centralSuccess

For every real ggg, let μg\mu_gμg​ be complex-plane volume weighted by the extended nonnegative density max⁡{1−gπgexp⁡(−1−gg∣z∣2),0}\max\{\frac{1-g}{\pi g}\exp(-\frac{1-g}{g}|z|^2),0\}max{πg1−g​exp(−g1−g​∣z∣2),0}, using zero for division by zero. The defined real number is the real conversion of

μg ⁣({z∈C:∣Re⁡z∣≤π/8 and ∣Im⁡z∣≤π/8}).\mu_g\!\left(\left\{z\in\mathbb C:|\operatorname{Re}z|\le\sqrt{\pi/8}\ \text{and}\ |\operatorname{Im}z|\le\sqrt{\pi/8}\right\}\right).μg​({z∈C:∣Rez∣≤π/8​ and ∣Imz∣≤π/8​}).

The square includes its boundary. The conversion sends an infinite measure value to zero, and no assumption 0<g<10<g<10<g<1 is made.

QAlgorithms.Papers.Independent.Intakenext100.P65.codeMoment

Given any pair of complex sequences B0,B1:N→CB_0,B_1:\mathbb N\to\mathbb CB0​,B1​:N→C and any nonnegative integer kkk, this defines

12(∑n=0∞nk∣(B0)n∣2+∑n=0∞nk∣(B1)n∣2).\frac12\left( \sum_{n=0}^{\infty}n^k|(B_0)_n|^2+ \sum_{n=0}^{\infty}n^k|(B_1)_n|^2 \right).21​(n=0∑∞​nk∣(B0​)n​∣2+n=0∑∞​nk∣(B1​)n​∣2).

The sequences need not be normalized, orthogonal, or square summable. Each infinite sum is evaluated separately and returns 000 if its series is not summable. For k=0k=0k=0, 00=10^0=100=1.

QAlgorithms.Papers.Independent.Intakenext100.P65.coherentCoordinates

For every real number aaa, this defines a complex sequence indexed by n∈Nn\in\mathbb Nn∈N by

(coherentCoordinates⁡(a))n=e−a2/2ann!,(\operatorname{coherentCoordinates}(a))_n =e^{-a^2/2}\frac{a^n}{\sqrt{n!}},(coherentCoordinates(a))n​=e−a2/2n!​an​,

with the real quantities embedded in C\mathbb CC. The parameter aaa is unrestricted, including negative values and zero; at a=0a=0a=0, the sequence has value 111 at n=0n=0n=0 and 000 elsewhere.

QAlgorithms.Papers.Independent.Intakenext100.P65.collectiveShift

For natural numbers d,Md,Md,M and a hypothesis d>0d>0d>0, this defines a complex matrix TTT, with rows and columns indexed by words in {0,…,d−1}M\{0,\ldots,d-1\}^{M}{0,…,d−1}M, by

Tx,y=1M∑k=0M−11 ⁣{yi={(xk+1) mod d,i=k,xi,i≠k for every i}.T_{x,y}=\frac1M\sum_{k=0}^{M-1} \mathbf1\!\left\{ y_i= \begin{cases} (x_k+1)\bmod d,&i=k,\\ x_i,&i\ne k \end{cases} \text{ for every }i \right\}.Tx,y​=M1​k=0∑M−1​1{yi​={(xk​+1)modd,xi​,​i=k,i=k​ for every i}.

Each coordinate contributes separately, even if several coordinates produce the same word. When M=0M=0M=0, the sum is empty and the reciprocal of MMM is zero, so the resulting matrix on the singleton set of empty words is zero. When d=1d=1d=1 and M>0M>0M>0, its sole entry is 111.

QAlgorithms.Papers.Independent.Intakenext100.P65.compressedShift

For natural numbers d,Md,Md,M and a hypothesis d>0d>0d>0, put N=(d−1)MN=(d-1)MN=(d−1)M. For r∈{0,…,N}r\in\{0,\ldots,N\}r∈{0,…,N}, let crc_rcr​ be the number of words x∈{0,…,d−1}Mx\in\{0,\ldots,d-1\}^{M}x∈{0,…,d−1}M with total excitation ∑ixi=r\sum_i x_i=r∑i​xi​=r, and define

Vx,r={1/cr,∑ixi=r,0,otherwise.V_{x,r}= \begin{cases} 1/\sqrt{c_r},&\sum_i x_i=r,\\ 0,&\text{otherwise}. \end{cases}Vx,r​={1/cr​​,0,​∑i​xi​=r,otherwise.​

This defines the (N+1)×(N+1)(N+1)\times(N+1)(N+1)×(N+1) complex matrix V†TVV^\dagger TVV†TV, where

Tx,y=1M∑k=0M−11{y equals x with coordinate k replaced by (xk+1) mod d}.T_{x,y}=\frac1M\sum_{k=0}^{M-1} \mathbf1\{y\text{ equals }x\text{ with coordinate }k \text{ replaced by }(x_k+1)\bmod d\}.Tx,y​=M1​k=0∑M−1​1{y equals x with coordinate k replaced by (xk​+1)modd}.

Reciprocals of zero are zero. In particular, M=0M=0M=0 is allowed and gives a zero 1×11\times11×1 matrix.

QAlgorithms.Papers.Independent.Intakenext100.P65.correctableCell

For either shape s∈{square,hexagonal}s\in\{\mathrm{square},\mathrm{hexagonal}\}s∈{square,hexagonal}, let Lsquare(a,b)=π/2(a+ib)L_{\mathrm{square}}(a,b)=\sqrt{\pi/2}(a+ib)Lsquare​(a,b)=π/2​(a+ib) and Lhexagonal(a,b)=π/3(a+b(1/2+i3/2))L_{\mathrm{hexagonal}}(a,b)=\sqrt{\pi/\sqrt3}\bigl(a+b(1/2+i\sqrt3/2)\bigr)Lhexagonal​(a,b)=π/3​​(a+b(1/2+i3​/2)) for integers a,ba,ba,b. The defined set is {z∈C:∣z∣≤∣z−Ls(n)∣ for every n∈Z2}\{z\in\mathbb C:\lvert z\rvert\le\lvert z-L_s(n)\rvert\text{ for every }n\in\mathbb Z^2\}{z∈C:∣z∣≤∣z−Ls​(n)∣ for every n∈Z2}. Thus it includes all points whose distance to the origin is no greater than their distance to any lattice point, including points where equality holds.

QAlgorithms.Papers.Independent.Intakenext100.P65.cosetPauli

For every c=(a,b)∈{0,1}2c=(a,b)\in\{0,1\}^2c=(a,b)∈{0,1}2, the defined complex 2×22\times22×2 matrix is XaZbX^aZ^bXaZb, where

X=(0110),Z=(100−1).X=\begin{pmatrix}0&1\\1&0\end{pmatrix}, \qquad Z=\begin{pmatrix}1&0\\0&-1\end{pmatrix}.X=(01​10​),Z=(10​0−1​).

The order is XXX followed by ZZZ in the matrix product: the four outputs for (a,b)=(0,0),(1,0),(0,1),(1,1)(a,b)=(0,0),(1,0),(0,1),(1,1)(a,b)=(0,0),(1,0),(0,1),(1,1) are respectively I,X,Z,XZI,X,Z,XZI,X,Z,XZ.

QAlgorithms.Papers.Independent.Intakenext100.P65.cosetProbability

For either shape sss, every real ggg, and every c∈{0,1}2c\in\{0,1\}^2c∈{0,1}2, let Lsquare(a,b)=π/2(a+ib)L_{\mathrm{square}}(a,b)=\sqrt{\pi/2}(a+ib)Lsquare​(a,b)=π/2​(a+ib) and Lhexagonal(a,b)=π/3(a+b(1/2+i3/2))L_{\mathrm{hexagonal}}(a,b)=\sqrt{\pi/\sqrt3}\bigl(a+b(1/2+i\sqrt3/2)\bigr)Lhexagonal​(a,b)=π/3​​(a+b(1/2+i3​/2)). Classically choose ns(z)∈Z2n_s(z)\in\mathbb Z^2ns​(z)∈Z2 minimizing ∣z−Ls(n)∣\lvert z-L_s(n)\rvert∣z−Ls​(n)∣, with unspecified tie resolution and an unspecified pair if no minimizer exists, and let Cs(z)C_s(z)Cs​(z) be its coordinatewise integer remainder modulo 222. Let μg\mu_gμg​ be complex-plane volume weighted by max⁡(dg,0)\max(d_g,0)max(dg​,0), where dg(z)=1−gπgexp⁡(−1−gg∣z∣2)d_g(z)=\frac{1-g}{\pi g}\exp(-\frac{1-g}{g}|z|^2)dg​(z)=πg1−g​exp(−g1−g​∣z∣2), using zero for division by zero. The defined real number is μg({z:Cs(z)=c})\mu_g(\{z:C_s(z)=c\})μg​({z:Cs​(z)=c}) converted from the extended nonnegative reals to the reals; this conversion sends an infinite measure value to zero. No restriction on ggg or measurability assertion about this event is included.

QAlgorithms.Papers.Independent.Intakenext100.P65.displaceDensity

For any z∈Cz\in\mathbb Cz∈C, any complex array R:N×N→CR:\mathbb N\times\mathbb N\to\mathbb CR:N×N→C, and any n,m∈Nn,m\in\mathbb Nn,m∈N, define

(DzR)nm=∑j∈N(∑k∈NDz(n,j)RjkDz(m,k)‾),(\mathcal D_zR)_{nm} =\sum_{j\in\mathbb N}\left( \sum_{k\in\mathbb N} D_z(n,j)R_{jk}\overline{D_z(m,k)} \right),(Dz​R)nm​=j∈N∑​(k∈N∑​Dz​(n,j)Rjk​Dz​(m,k)​),

where, for a,b∈Na,b\in\mathbb Na,b∈N,

Dz(a,b)=e−∣z∣2/2a! b!∑h=0min⁡(a,b)za−h(−z‾)b−h(a−h)! (b−h)! h!.D_z(a,b) =e^{-|z|^2/2}\sqrt{a!\,b!} \sum_{h=0}^{\min(a,b)} \frac{z^{a-h}(-\overline z)^{b-h}} {(a-h)!\,(b-h)!\,h!}.Dz​(a,b)=e−∣z∣2/2a!b!​h=0∑min(a,b)​(a−h)!(b−h)!h!za−h(−z)b−h​.

The real prefactors are embedded in C\mathbb CC. No density or convergence assumption is imposed on RRR; the inner sum is taken first, and each infinite sum is assigned zero if its summands are not summable.

QAlgorithms.Papers.Independent.Intakenext100.P65.displaceWave

For every complex number zzz, every function v:R→Cv:\mathbb R\to\mathbb Cv:R→C, and every x∈Rx\in\mathbb Rx∈R, define the displaced function by

(Dzv)(x)=exp⁡ ⁣(i(2 Im⁡(z)x−Re⁡(z)Im⁡(z)))v ⁣(x−2 Re⁡(z)).(D_zv)(x)=\exp\!\left(i\bigl(\sqrt2\,\operatorname{Im}(z)x-\operatorname{Re}(z)\operatorname{Im}(z)\bigr)\right)v\!\left(x-\sqrt2\,\operatorname{Re}(z)\right).(Dz​v)(x)=exp(i(2​Im(z)x−Re(z)Im(z)))v(x−2​Re(z)).

No integrability, continuity, or normalization condition is imposed on vvv.

QAlgorithms.Papers.Independent.Intakenext100.P65.embedDicke

For natural numbers M,S,KM,S,KM,S,K, complex coefficients v0,…,vMv_0,\ldots,v_Mv0​,…,vM​, and a configuration x:{0,…,K−1}→{0,1}x:\{0,\ldots,K-1\}\to\{0,1\}x:{0,…,K−1}→{0,1}, this defines

embedDicke⁡M,S,K(v)(x)=∑m=0Mvm{1(K(S+1)m),#{i:xi=1}=(S+1)m,0,otherwise.\operatorname{embedDicke}_{M,S,K}(v)(x) =\sum_{m=0}^{M}v_m \begin{cases} \displaystyle\frac{1}{\sqrt{\binom K{(S+1)m}}},& \#\{i:x_i=1\}=(S+1)m,\\ 0,&\text{otherwise}. \end{cases}embedDickeM,S,K​(v)(x)=m=0∑M​vm​⎩⎨⎧​((S+1)mK​)​1​,0,​#{i:xi​=1}=(S+1)m,otherwise.​

The real coefficients are regarded as complex numbers. All parameters may be zero. If (S+1)m>K(S+1)m>K(S+1)m>K, that summand is zero for every xxx; the inverse of zero used in the definition is also assigned the value zero.

QAlgorithms.Papers.Independent.Intakenext100.P65.embedOne

For every natural numbers M,SM,SM,S and arbitrary complex vector v=(v0,…,vM)v=(v_0,\ldots,v_M)v=(v0​,…,vM​), this defines the complex-valued sequence

u(n)=∑m=0M{vm,n=(S+1)m,0,otherwise.u(n)=\sum_{m=0}^{M} \begin{cases} v_m,&n=(S+1)m,\\ 0,&\text{otherwise}. \end{cases}u(n)=m=0∑M​{vm​,0,​n=(S+1)m,otherwise.​

Since S+1>0S+1>0S+1>0, equivalently u((S+1)m)=vmu((S+1)m)=v_mu((S+1)m)=vm​ for 0≤m≤M0\le m\le M0≤m≤M, with zero at every other coordinate. There is no normalization assumption on vvv; S=0S=0S=0 places its entries at coordinates 0,…,M0,\ldots,M0,…,M, and M=0M=0M=0 places only v0v_0v0​ at coordinate 000.

QAlgorithms.Papers.Independent.Intakenext100.P65.embedTwo

For every natural numbers M,SM,SM,S and arbitrary complex vector v=(v0,…,vM)v=(v_0,\ldots,v_M)v=(v0​,…,vM​), this defines a function u:N×N→Cu:\mathbb N\times\mathbb N\to\mathbb Cu:N×N→C by

u(n1,n2)=∑m=0M{vm,(n1,n2)=((S+1)(M−m),(S+1)m),0,otherwise.u(n_1,n_2)=\sum_{m=0}^{M} \begin{cases} v_m,&(n_1,n_2)=((S+1)(M-m),(S+1)m),\\ 0,&\text{otherwise}. \end{cases}u(n1​,n2​)=m=0∑M​{vm​,0,​(n1​,n2​)=((S+1)(M−m),(S+1)m),otherwise.​

The subtraction M−mM-mM−m is natural-number subtraction, with m≤Mm\le Mm≤M guaranteed by its index range. No normalization is required. The definition includes S=0S=0S=0 and M=0M=0M=0; in the latter case its only potentially nonzero value is u(0,0)=v0u(0,0)=v_0u(0,0)=v0​.

QAlgorithms.Papers.Independent.Intakenext100.P65.extendedCoefficients

For natural numbers d,Md,Md,M, an index 0≤μ<d0\le\mu<d0≤μ<d, and an index 0≤m<(d−1)M+10\le m<(d-1)M+10≤m<(d−1)M+1, this defines

cd,M,μ(m)=exp⁡ ⁣(i 2πμmd)Cd,M(m)dM,Cd,M(m)=#{x∈{0,…,d−1}M:∑ixi=m}.c_{d,M,\mu}(m)= \exp\!\left(i\,\frac{2\pi\mu m}{d}\right) \frac{\sqrt{C_{d,M}(m)}}{\sqrt{d^M}}, \qquad C_{d,M}(m)= \#\left\{x\in\{0,\ldots,d-1\}^{M}:\sum_i x_i=m\right\}.cd,M,μ​(m)=exp(id2πμm​)dM​Cd,M​(m)​​,Cd,M​(m)=#{x∈{0,…,d−1}M:i∑​xi​=m}.

The square-root ratio is real and then embedded into C\mathbb CC; the subtraction d−1d-1d−1 is natural-number subtraction, truncated at zero. Although ddd is quantified over all natural numbers, an index μ\muμ exists only when d>0d>0d>0. When M=0M=0M=0, the sole index is m=0m=0m=0 and its coefficient is 111.

QAlgorithms.Papers.Independent.Intakenext100.P65.extendedCount

For arbitrary natural numbers d,M,md,M,md,M,m, this defines

Cd,M(m)=#{x∈{0,…,d−1}M:∑i=0M−1xi=m}.C_{d,M}(m)= \#\left\{x\in\{0,\ldots,d-1\}^{M}: \sum_{i=0}^{M-1}x_i=m\right\}.Cd,M​(m)=#{x∈{0,…,d−1}M:i=0∑M−1​xi​=m}.

In particular, Cd,0(0)=1C_{d,0}(0)=1Cd,0​(0)=1 and Cd,0(m)=0C_{d,0}(m)=0Cd,0​(m)=0 for m>0m>0m>0; for d=0d=0d=0 and M>0M>0M>0, every count is zero. For d≥1d\ge1d≥1, the count is zero when m>(d−1)Mm>(d-1)Mm>(d−1)M.

QAlgorithms.Papers.Independent.Intakenext100.P65.extendedVector

For arbitrary natural numbers d,M,md,M,md,M,m and a configuration x∈{0,…,d−1}Mx\in\{0,\ldots,d-1\}^{M}x∈{0,…,d−1}M, this defines the complex amplitude

Vd,M,m(x)={1Cd,M(m),∑i=0M−1xi=m,0,otherwise,Cd,M(m)=#{y∈{0,…,d−1}M:∑iyi=m}.V_{d,M,m}(x)= \begin{cases} \displaystyle\frac1{\sqrt{C_{d,M}(m)}},& \sum_{i=0}^{M-1}x_i=m,\\ 0,&\text{otherwise}, \end{cases} \qquad C_{d,M}(m)= \#\left\{y\in\{0,\ldots,d-1\}^{M}:\sum_i y_i=m\right\}.Vd,M,m​(x)=⎩⎨⎧​Cd,M​(m)​1​,0,​∑i=0M−1​xi​=m,otherwise,​Cd,M​(m)=#{y∈{0,…,d−1}M:i∑​yi​=m}.

The real value is embedded into C\mathbb CC. If the count is zero, the function is zero; division by zero is defined to return zero. For M=0M=0M=0, its value on the empty configuration is 111 when m=0m=0m=0 and 000 otherwise.

QAlgorithms.Papers.Independent.Intakenext100.P65.fockDensity

For an arbitrary array R:N×N→CR:\mathbb N\times\mathbb N\to\mathbb CR:N×N→C, this defines the following conjunction: for every natural number ddd, the finite matrix (Rij)0≤i,j<d(R_{ij})_{0\le i,j<d}(Rij​)0≤i,j<d​ is positive semidefinite; the real sequence n↦Re⁡Rnnn\mapsto\operatorname{Re}R_{nn}n↦ReRnn​ is summable; and

∑n=0∞Re⁡Rnn=1.\sum_{n=0}^{\infty}\operatorname{Re}R_{nn}=1.n=0∑∞​ReRnn​=1.

The finite-matrix condition includes d=0d=0d=0, where it is vacuous. No separate trace-class hypothesis is included.

QAlgorithms.Papers.Independent.Intakenext100.P65.fockDomain

For any complex sequence v:N→Cv:\mathbb N\to\mathbb Cv:N→C, the predicate holds precisely when the real series ∑n=0∞∣vn∣2\sum_{n=0}^{\infty}|v_n|^2∑n=0∞​∣vn​∣2 is summable. It does not require the sum to equal 111; in particular, the zero sequence satisfies it.

QAlgorithms.Papers.Independent.Intakenext100.P65.fockInner

For any two complex sequences v,w:N→Cv,w:\mathbb N\to\mathbb Cv,w:N→C, this defines

∑n=0∞vn‾ wn∈C.\sum_{n=0}^{\infty}\overline{v_n}\,w_n\in\mathbb C.n=0∑∞​vn​​wn​∈C.

Neither sequence is required to be square summable or normalized. If the displayed complex series is not summable, the total infinite-sum operation returns 000.

QAlgorithms.Papers.Independent.Intakenext100.P65.fockMomentDomain

For every complex sequence v:N→Cv:\mathbb N\to\mathbb Cv:N→C and every nonnegative integer kkk, this predicate means that the real series

∑n=0∞nk∣vn∣2\sum_{n=0}^{\infty}n^k|v_n|^2n=0∑∞​nk∣vn​∣2

is summable. For k=0k=0k=0, including the convention 00=10^0=100=1, this is exactly square summability of vvv. For k>0k>0k>0, the term at n=0n=0n=0 is zero. No normalization is required.

QAlgorithms.Papers.Independent.Intakenext100.P65.fockNormSq

For any complex sequence v:N→Cv:\mathbb N\to\mathbb Cv:N→C, this defines the real number

∑n=0∞∣vn∣2.\sum_{n=0}^{\infty}|v_n|^2.n=0∑∞​∣vn​∣2.

No summability hypothesis is required. If this real series is not summable, the total infinite-sum operation used here returns 000.

QAlgorithms.Papers.Independent.Intakenext100.P65.fourierProduct

For natural numbers d,Md,Md,M, a label μ∈{0,…,d−1}\mu\in\{0,\ldots,d-1\}μ∈{0,…,d−1}, and a word x:{0,…,M−1}→{0,…,d−1}x:\{0,\ldots,M-1\}\to\{0,\ldots,d-1\}x:{0,…,M−1}→{0,…,d−1}, this defines the complex number

Fd,M,μ(x)=∏i=0M−1exp⁡ ⁣(2πiμxi/d)d.F_{d,M,\mu}(x)= \prod_{i=0}^{M-1} \frac{\exp\!\left(2\pi\mathrm{i}\mu x_i/d\right)}{\sqrt d}.Fd,M,μ​(x)=i=0∏M−1​d​exp(2πiμxi​/d)​.

For M=0M=0M=0, the empty product is 111. For d=0d=0d=0, there is no possible label μ\muμ, so the definition has no instances with that value of ddd.

QAlgorithms.Papers.Independent.Intakenext100.P65.gkpFock

For every real ddd and μ∈{0,1}\mu\in\{0,1\}μ∈{0,1}, define a sequence cd,μ:N→Cc_{d,\mu}:\mathbb N\to\mathbb Ccd,μ​:N→C by

cd,μ(n)=∫Rhn(x)Wμ,d(x)Nμ,d dx,hn(x)=Hn(x)e−x2/22nn!π,c_{d,\mu}(n)=\int_{\mathbb R}h_n(x)\frac{W_{\mu,d}(x)}{\sqrt{N_{\mu,d}}}\,dx, \qquad h_n(x)=\frac{H_n(x)e^{-x^2/2}}{\sqrt{2^n n!\sqrt\pi}},cd,μ​(n)=∫R​hn​(x)Nμ,d​​Wμ,d​(x)​dx,hn​(x)=2nn!π​​Hn​(x)e−x2/2​,

where H0(X)=1H_0(X)=1H0​(X)=1, Hn+1(X)=2XHn(X)−Hn′(X)H_{n+1}(X)=2XH_n(X)-H_n'(X)Hn+1​(X)=2XHn​(X)−Hn′​(X), and

Wμ,d(x)=2π∑r∈Ze−(x−π(2r+μ))2/(2d2)−d2(x+π(2r+μ))2/8,Nμ,d=∫R∣Wμ,d(x)∣2 dx.W_{\mu,d}(x)=\sqrt{\frac2{\sqrt\pi}}\sum_{r\in\mathbb Z} e^{-(x-\sqrt\pi(2r+\mu))^2/(2d^2)-d^2(x+\sqrt\pi(2r+\mu))^2/8}, \qquad N_{\mu,d}=\int_{\mathbb R}|W_{\mu,d}(x)|^2\,dx.Wμ,d​(x)=π​2​​r∈Z∑​e−(x−π​(2r+μ))2/(2d2)−d2(x+π​(2r+μ))2/8,Nμ,d​=∫R​∣Wμ,d​(x)∣2dx.

Real factors are regarded as complex when necessary. This defines coordinates for all n≥0n\ge0n≥0 without asserting square summability or normalization, and it includes d=0d=0d=0: division by zero gives zero, nonsummable series have sum zero, and nonintegrable functions have Lebesgue integral zero.

QAlgorithms.Papers.Independent.Intakenext100.P65.gkpNormalizerPoint

For either shape square\mathrm{square}square or hexagonal\mathrm{hexagonal}hexagonal and every integer pair (r,s)∈Z2(r,s)\in\mathbb Z^2(r,s)∈Z2, define a complex point by

Lsquare(r,s)=π2(r+is),Lhexagonal(r,s)=π3(r+s(12+i32)).L_{\mathrm{square}}(r,s)=\sqrt{\frac\pi2}(r+is),\qquad L_{\mathrm{hexagonal}}(r,s)= \sqrt{\frac\pi{\sqrt3}}\left(r+s\left(\frac12+i\frac{\sqrt3}{2}\right)\right).Lsquare​(r,s)=2π​​(r+is),Lhexagonal​(r,s)=3​π​​(r+s(21​+i23​​)).

Both integer coordinates may be positive, negative, or zero; in particular, either shape sends (0,0)(0,0)(0,0) to 000.

QAlgorithms.Papers.Independent.Intakenext100.P65.hermitePhys

Define a sequence of real polynomials HnH_nHn​ for every natural number nnn, including zero, recursively by

H0(X)=1,Hn+1(X)=2XHn(X)−Hn′(X).H_0(X)=1,\qquad H_{n+1}(X)=2XH_n(X)-H_n'(X).H0​(X)=1,Hn+1​(X)=2XHn​(X)−Hn′​(X).

QAlgorithms.Papers.Independent.Intakenext100.P65.hermiteWave

For every natural number nnn, including n=0n=0n=0, and every real xxx, define the real-valued function

hn(x)=Hn(x)e−x2/22nn!π,h_n(x)=\frac{H_n(x)e^{-x^2/2}}{\sqrt{2^n n!\sqrt\pi}},hn​(x)=2nn!π​​Hn​(x)e−x2/2​,

where the real polynomials satisfy H0(X)=1H_0(X)=1H0​(X)=1 and Hn+1(X)=2XHn(X)−Hn′(X)H_{n+1}(X)=2XH_n(X)-H_n'(X)Hn+1​(X)=2XHn​(X)−Hn′​(X). In particular, h0(x)=π−1/4e−x2/2h_0(x)=\pi^{-1/4}e^{-x^2/2}h0​(x)=π−1/4e−x2/2.

QAlgorithms.Papers.Independent.Intakenext100.P65.idealComb

For either μ∈{0,1}\mu\in\{0,1\}μ∈{0,1} and any function f:R→Cf:\mathbb R\to\mathbb Cf:R→C, define

Iμ(f)=∑n∈Zf ⁣(π(2n+μ)).I_\mu(f)=\sum_{n\in\mathbb Z} f\!\left(\sqrt\pi(2n+\mu)\right).Iμ​(f)=n∈Z∑​f(π​(2n+μ)).

No regularity or summability assumption is imposed on fff; this sum is assigned zero if the family of sampled values is not summable.

QAlgorithms.Papers.Independent.Intakenext100.P65.idealDisplacementKernel

For any μ,ν∈{0,1}\mu,\nu\in\{0,1\}μ,ν∈{0,1} and any function f:C→Cf:\mathbb C\to\mathbb Cf:C→C, define

Kμν(f)=∑(a,b)∈Z2exp⁡ ⁣(iπ(a+μ+ν2)b)f ⁣(π2(2a+μ−ν+ib)).K_{\mu\nu}(f) =\sum_{(a,b)\in\mathbb Z^2} \exp\!\left(i\pi\left(a+\frac{\mu+\nu}{2}\right)b\right) f\!\left(\sqrt{\frac{\pi}{2}}(2a+\mu-\nu+ib)\right).Kμν​(f)=(a,b)∈Z2∑​exp(iπ(a+2μ+ν​)b)f(2π​​(2a+μ−ν+ib)).

No regularity or summability assumption is imposed on fff. This is a sum over integer pairs as a single family, and it equals zero if that family is not summable.

QAlgorithms.Papers.Independent.Intakenext100.P65.idealRecoveredChannel

For either shape sss and every real ggg, use the lattice maps Lsquare(a,b)=π/2(a+ib)L_{\mathrm{square}}(a,b)=\sqrt{\pi/2}(a+ib)Lsquare​(a,b)=π/2​(a+ib) and Lhexagonal(a,b)=π/3(a+b(1/2+i3/2))L_{\mathrm{hexagonal}}(a,b)=\sqrt{\pi/\sqrt3}\bigl(a+b(1/2+i\sqrt3/2)\bigr)Lhexagonal​(a,b)=π/3​​(a+b(1/2+i3​/2)). Let Cs(z)∈{0,1}2C_s(z)\in\{0,1\}^2Cs​(z)∈{0,1}2 be the coordinatewise parity of a classically selected nearest integer lattice label, with unspecified tie resolution and an unspecified label if no minimizer exists. Define pcp_cpc​ as the real conversion of the measure of {z:Cs(z)=c}\{z:C_s(z)=c\}{z:Cs​(z)=c} under complex-plane volume weighted by max⁡{1−gπgexp⁡(−1−gg∣z∣2),0}\max\{\frac{1-g}{\pi g}\exp(-\frac{1-g}{g}|z|^2),0\}max{πg1−g​exp(−g1−g​∣z∣2),0}; division by zero is zero, and an infinite measure converts to zero. With X=(0110)X=\begin{pmatrix}0&1\\1&0\end{pmatrix}X=(01​10​), Z=diag⁡(1,−1)Z=\operatorname{diag}(1,-1)Z=diag(1,−1), and Ka,b=p(a,b)XaZbK_{a,b}=\sqrt{p_{(a,b)}}X^aZ^bKa,b​=p(a,b)​​XaZb, the defined map on every complex 2×22\times22×2 matrix ρ\rhoρ is

Φs,g(ρ)=∑a,b∈{0,1}Ka,bρKa,b†.\Phi_{s,g}(\rho)=\sum_{a,b\in\{0,1\}}K_{a,b}\rho K_{a,b}^{\dagger}.Φs,g​(ρ)=a,b∈{0,1}∑​Ka,b​ρKa,b†​.

The definition itself imposes no restriction on ggg and does not assert the normalization ∑a,bKa,b†Ka,b=I\sum_{a,b}K_{a,b}^{\dagger}K_{a,b}=I∑a,b​Ka,b†​Ka,b​=I.

QAlgorithms.Papers.Independent.Intakenext100.P65.jumpTrajectory

For any real numbers k,tk,tk,t, natural number lll, list of real times τ0,…,τl−1\tau_0,\ldots,\tau_{l-1}τ0​,…,τl−1​, and complex sequence vvv, start with s0=0s_0=0s0​=0 and w(0)=vw^{(0)}=vw(0)=v. For j=0,…,l−1j=0,\ldots,l-1j=0,…,l−1, set sj+1=τjs_{j+1}=\tau_jsj+1​=τj​ and

wn(j+1)=n+1 e−k(τj−sj)(n+1)/2wn+1(j)(n∈N).w^{(j+1)}_n=\sqrt{n+1}\,e^{-k(\tau_j-s_j)(n+1)/2}w^{(j)}_{n+1}\qquad(n\in\mathbb N).wn(j+1)​=n+1​e−k(τj​−sj​)(n+1)/2wn+1(j)​(n∈N).

The defined output has coordinates e−k(t−sl)n/2wn(l)e^{-k(t-s_l)n/2}w^{(l)}_ne−k(t−sl​)n/2wn(l)​. Thus the times are processed in index order, with no assumption that they increase or lie between 000 and ttt; repeated times are allowed. No summability or normalization is required. When l=0l=0l=0, the output is n↦e−ktn/2vnn\mapsto e^{-ktn/2}v_nn↦e−ktn/2vn​.

QAlgorithms.Papers.Independent.Intakenext100.P65.kernelLattice

For any μ,ν∈{0,1}\mu,\nu\in\{0,1\}μ,ν∈{0,1} and any integer pair (a,b)∈Z2(a,b)\in\mathbb Z^2(a,b)∈Z2, define the complex number

Λμν(a,b)=π2(2a+μ−ν+ib),\Lambda_{\mu\nu}(a,b) =\sqrt{\frac{\pi}{2}}\bigl(2a+\mu-\nu+ib\bigr),Λμν​(a,b)=2π​​(2a+μ−ν+ib),

where the integer labels and real square root are embedded in C\mathbb CC.

QAlgorithms.Papers.Independent.Intakenext100.P65.kernelPhase

For any μ,ν∈{0,1}\mu,\nu\in\{0,1\}μ,ν∈{0,1} and any integer pair (a,b)∈Z2(a,b)\in\mathbb Z^2(a,b)∈Z2, define

χμν(a,b)=exp⁡ ⁣(iπ(a+μ+ν2)b).\chi_{\mu\nu}(a,b) =\exp\!\left(i\pi\left(a+\frac{\mu+\nu}{2}\right)b\right).χμν​(a,b)=exp(iπ(a+2μ+ν​)b).

QAlgorithms.Papers.Independent.Intakenext100.P65.liouville

For any function TTT from complex 2×22\times22×2 matrices to complex 2×22\times22×2 matrices, liouville(T)\mathrm{liouville}(T)liouville(T) is the complex 4×44\times44×4 matrix with entries

liouville(T)kl=12tr⁡ ⁣(PkT(Pl)),k,l∈{0,1,2,3},\mathrm{liouville}(T)_{kl}=\frac12\operatorname{tr}\!\bigl(P_kT(P_l)\bigr), \qquad k,l\in\{0,1,2,3\},liouville(T)kl​=21​tr(Pk​T(Pl​)),k,l∈{0,1,2,3},

where P0=I2P_0=I_2P0​=I2​, P1=(0110)P_1=\left(\begin{smallmatrix}0&1\\1&0\end{smallmatrix}\right)P1​=(01​10​), P2=(0−ii0)P_2=\left(\begin{smallmatrix}0&-\mathrm i\\\mathrm i&0\end{smallmatrix}\right)P2​=(0i​−i0​), and P3=diag⁡(1,−1)P_3=\operatorname{diag}(1,-1)P3​=diag(1,−1). No linearity, positivity, or trace preservation is required of TTT, and the entries retain their complex values.

QAlgorithms.Papers.Independent.Intakenext100.P65.liouvilleBasis

The function liouvilleBasis\mathrm{liouvilleBasis}liouvilleBasis, indexed by 0,1,2,30,1,2,30,1,2,3, returns the following complex 2×22\times22×2 matrices, whose rows and columns are ordered 0,10,10,1:

P0=(1001),P1=(0110),P2=(0−ii0),P3=(100−1).P_0=\begin{pmatrix}1&0\\0&1\end{pmatrix},\qquad P_1=\begin{pmatrix}0&1\\1&0\end{pmatrix},\qquad P_2=\begin{pmatrix}0&-\mathrm i\\ \mathrm i&0\end{pmatrix},\qquad P_3=\begin{pmatrix}1&0\\0&-1\end{pmatrix}.P0​=(10​01​),P1​=(01​10​),P2​=(0i​−i0​),P3​=(10​0−1​).

QAlgorithms.Papers.Independent.Intakenext100.P65.logicalCoset

For either shape s∈{square,hexagonal}s\in\{\mathrm{square},\mathrm{hexagonal}\}s∈{square,hexagonal} and every z∈Cz\in\mathbb Cz∈C, define Lsquare(a,b)=π/2(a+ib)L_{\mathrm{square}}(a,b)=\sqrt{\pi/2}(a+ib)Lsquare​(a,b)=π/2​(a+ib) and Lhexagonal(a,b)=π/3(a+b(1/2+i3/2))L_{\mathrm{hexagonal}}(a,b)=\sqrt{\pi/\sqrt3}\bigl(a+b(1/2+i\sqrt3/2)\bigr)Lhexagonal​(a,b)=π/3​​(a+b(1/2+i3​/2)). Let (a,b)(a,b)(a,b) be the classically selected integer pair minimizing the distance from zzz to Ls(a,b)L_s(a,b)Ls​(a,b), with unspecified tie resolution and an unspecified pair if there is no minimizer. The output is (a mod 2,b mod 2)(a\bmod2,b\bmod2)(amod2,bmod2), regarded as a pair in {0,1}2\{0,1\}^2{0,1}2. Integer remainders modulo the positive integer 222 are used, so negative coordinates also give remainders 000 or 111.

QAlgorithms.Papers.Independent.Intakenext100.P65.lossChannel

For any real ggg, any complex array R:N×N→CR:\mathbb N\times\mathbb N\to\mathbb CR:N×N→C, and any n,m∈Nn,m\in\mathbb Nn,m∈N, define

(LgR)nm=∑l∈Ncg(l,n+l)cg(l,m+l)Rn+l,m+l,cg(l,s)={(sl) (g)l(1−g)s−l,l≤s,0,l>s.(\mathcal L_gR)_{nm} =\sum_{l\in\mathbb N} c_g(l,n+l)c_g(l,m+l)R_{n+l,m+l}, \qquad c_g(l,s)= \begin{cases} \sqrt{\binom{s}{l}}\,(\sqrt g)^l(\sqrt{1-g})^{s-l},&l\le s,\\ 0,&l>s. \end{cases}(Lg​R)nm​=l∈N∑​cg​(l,n+l)cg​(l,m+l)Rn+l,m+l​,cg​(l,s)={(ls​)​(g​)l(1−g​)s−l,0,​l≤s,l>s.​

The real coefficients are regarded as complex scalars. There is no restriction on ggg, no positivity or normalization assumption on RRR, and no summability hypothesis; the infinite sum has value zero if its summands are not summable. Square roots are real square roots, which equal zero on negative inputs, and natural-number subtraction is truncated at zero.

QAlgorithms.Papers.Independent.Intakenext100.P65.lossCoefficient

For every real number ggg and nonnegative integers l,nl,nl,n, this defines

{(nl) (g)l(1−g)n−l,l≤n,0,l>n.\begin{cases} \displaystyle \sqrt{\binom nl}\,(\sqrt g)^l(\sqrt{1-g})^{n-l},&l\le n,\\ 0,&l>n. \end{cases}⎩⎨⎧​(ln​)​(g​)l(1−g​)n−l,0,​l≤n,l>n.​

There is no restriction 0≤g≤10\le g\le10≤g≤1. Real square roots of negative arguments are defined to be zero, and every zeroth power, including 000^000, equals 111. The subtraction n−ln-ln−l is natural-number subtraction, and the branch using it is selected only when l≤nl\le nl≤n.

QAlgorithms.Papers.Independent.Intakenext100.P65.lossDisplacementIntegrand

For any real ggg, natural numbers l,k,n,ml,k,n,ml,k,n,m, and complex number zzz, this defines

Ig,l,k,n,m(z)=e−(1−g)∣z∣2/2π Dz‾(l,k) Dg z(n,m),I_{g,l,k,n,m}(z)= \frac{e^{-(1-g)|z|^2/2}}{\pi}\, D_{\overline z}(l,k)\,D_{\sqrt g\,z}(n,m),Ig,l,k,n,m​(z)=πe−(1−g)∣z∣2/2​Dz​(l,k)Dg​z​(n,m),

where, for a complex number www and natural numbers r,sr,sr,s,

Dw(r,s)=e−∣w∣2/2r! s!∑h=0min⁡(r,s)wr−h(−w‾)s−h(r−h)! (s−h)! h!.D_w(r,s)= e^{-|w|^2/2}\sqrt{r!\,s!} \sum_{h=0}^{\min(r,s)} \frac{w^{r-h}(-\overline w)^{s-h}} {(r-h)!\,(s-h)!\,h!}.Dw​(r,s)=e−∣w∣2/2r!s!​h=0∑min(r,s)​(r−h)!(s−h)!h!wr−h(−w)s−h​.

Real factors are embedded into C\mathbb CC. The definition imposes no restriction on ggg; in particular, g=0\sqrt g=0g​=0 when g<0g<0g<0.

QAlgorithms.Papers.Independent.Intakenext100.P65.lossEntry

For any real ggg and natural numbers l,n,ml,n,ml,n,m, this defines

Lg(l;n,m)={(ml) (g)l(1−g)m−l,m=n+l,0,m≠n+l,L_g(l;n,m)= \begin{cases} \displaystyle \sqrt{\binom ml}\,(\sqrt g)^l(\sqrt{1-g})^{m-l}, &m=n+l,\\ 0,&m\ne n+l, \end{cases}Lg​(l;n,m)=⎩⎨⎧​(lm​)​(g​)l(1−g​)m−l,0,​m=n+l,m=n+l,​

as a complex number by embedding the displayed real coefficient into C\mathbb CC. The condition m=n+lm=n+lm=n+l ensures l≤ml\le ml≤m, so the zero branch of the underlying loss coefficient for l>ml>ml>m cannot occur in the first case. There is no assumption that g∈[0,1]g\in[0,1]g∈[0,1]; real square roots of negative arguments are interpreted as zero, and zeroth powers remain 111.

QAlgorithms.Papers.Independent.Intakenext100.P65.lossProduct

For any real ggg and natural numbers l,k,n,ml,k,n,ml,k,n,m, this defines

∑j=0∞(1{n=j+l}(nl) (g)l(1−g)n−l)‾(1{m=j+k}(mk) (g)k(1−g)m−k),\sum_{j=0}^{\infty} \overline{ \left( \mathbf1\{n=j+l\} \sqrt{\binom nl}\,(\sqrt g)^l(\sqrt{1-g})^{n-l} \right)} \left( \mathbf1\{m=j+k\} \sqrt{\binom mk}\,(\sqrt g)^k(\sqrt{1-g})^{m-k} \right),j=0∑∞​(1{n=j+l}(ln​)​(g​)l(1−g​)n−l)​(1{m=j+k}(km​)​(g​)k(1−g​)m−k),

where the real coefficients are regarded as complex numbers. At most one summand can be nonzero, and all summands vanish unless some natural number jjj satisfies both n=j+ln=j+ln=j+l and m=j+km=j+km=j+k. No restriction on ggg is imposed; natural-number subtraction is truncated at zero, and real square roots of negative arguments are zero.

QAlgorithms.Papers.Independent.Intakenext100.P65.lossVector

For every real number ggg, nonnegative integer lll, and complex sequence v:N→Cv:\mathbb N\to\mathbb Cv:N→C, this defines the complex sequence

n⟼(n+ll) (g)l(1−g)n vn+l,n≥0.n\longmapsto \sqrt{\binom{n+l}{l}}\,(\sqrt g)^l(\sqrt{1-g})^n\,v_{n+l}, \qquad n\ge0.n⟼(ln+l​)​(g​)l(1−g​)nvn+l​,n≥0.

The coefficient uses the nonzero-branch formula because l≤n+ll\le n+ll≤n+l always holds. No square-summability or normalization condition is imposed, and ggg may be any real number. Negative square-root arguments give zero, while zeroth powers, including 000^000, give 111.

QAlgorithms.Papers.Independent.Intakenext100.P65.modularSyndrome

For either shape s∈{square,hexagonal}s\in\{\mathrm{square},\mathrm{hexagonal}\}s∈{square,hexagonal} and every z∈Cz\in\mathbb Cz∈C, put Lsquare(a,b)=π/2(a+ib)L_{\mathrm{square}}(a,b)=\sqrt{\pi/2}(a+ib)Lsquare​(a,b)=π/2​(a+ib) and Lhexagonal(a,b)=π/3(a+b(1/2+i3/2))L_{\mathrm{hexagonal}}(a,b)=\sqrt{\pi/\sqrt3}\bigl(a+b(1/2+i\sqrt3/2)\bigr)Lhexagonal​(a,b)=π/3​​(a+b(1/2+i3​/2)). Let ns(z)∈Z2n_s(z)\in\mathbb Z^2ns​(z)∈Z2 be the classically selected pair minimizing ∣z−Ls(n)∣\lvert z-L_s(n)\rvert∣z−Ls​(n)∣ over integer pairs, with an unspecified choice among ties and an unspecified returned pair if no minimizer exists. The defined complex number is z−Ls(ns(z))z-L_s(n_s(z))z−Ls​(ns​(z)).

QAlgorithms.Papers.Independent.Intakenext100.P65.nearestLattice

For either shape s∈{square,hexagonal}s\in\{\mathrm{square},\mathrm{hexagonal}\}s∈{square,hexagonal} and every z∈Cz\in\mathbb Cz∈C, define the lattice-point map, for (a,b)∈Z2(a,b)\in\mathbb Z^2(a,b)∈Z2, by Lsquare(a,b)=π/2(a+ib)L_{\mathrm{square}}(a,b)=\sqrt{\pi/2}(a+ib)Lsquare​(a,b)=π/2​(a+ib) and Lhexagonal(a,b)=π/3(a+b(1/2+i3/2))L_{\mathrm{hexagonal}}(a,b)=\sqrt{\pi/\sqrt3}\bigl(a+b(1/2+i\sqrt3/2)\bigr)Lhexagonal​(a,b)=π/3​​(a+b(1/2+i3​/2)). This definition selects an integer pair nnn satisfying ∣z−Ls(n)∣≤∣z−Ls(m)∣\lvert z-L_s(n)\rvert\le\lvert z-L_s(m)\rvert∣z−Ls​(n)∣≤∣z−Ls​(m)∣ for every m∈Z2m\in\mathbb Z^2m∈Z2, if such a pair exists. The selection is by classical choice, with no specified rule for resolving ties; if no minimizing pair exists, it still returns an unspecified integer pair.

QAlgorithms.Papers.Independent.Intakenext100.P65.nextLabel

For a natural number ddd, a hypothesis d>0d>0d>0, and j∈{0,…,d−1}j\in\{0,\ldots,d-1\}j∈{0,…,d−1}, this defines the label (j+1) mod d(j+1)\bmod d(j+1)modd, regarded as an element of the same label set. Thus d−1d-1d−1 maps to 000; when d=1d=1d=1, the sole label maps to itself.

QAlgorithms.Papers.Independent.Intakenext100.P65.noJump

For any real numbers k,tk,tk,t and any complex sequence v:N→Cv:\mathbb N\to\mathbb Cv:N→C, this defines the complex sequence

(noJump⁡(k,t,v))n=e−ktn/2vn(n∈N).(\operatorname{noJump}(k,t,v))_n=e^{-ktn/2}v_n\qquad(n\in\mathbb N).(noJump(k,t,v))n​=e−ktn/2vn​(n∈N).

There are no positivity, normalization, or summability assumptions; kkk and ttt may be zero or negative.

QAlgorithms.Papers.Independent.Intakenext100.P65.normalizedSmearedWave

For μ∈{0,1}\mu\in\{0,1\}μ∈{0,1} and every d,x∈Rd,x\in\mathbb Rd,x∈R, define

W^μ,d(x)=Wμ,d(x)Nμ,d,Nμ,d=∫R∣Wμ,d(y)∣2 dy,\widehat W_{\mu,d}(x)=\frac{W_{\mu,d}(x)}{\sqrt{N_{\mu,d}}}, \qquad N_{\mu,d}=\int_{\mathbb R}|W_{\mu,d}(y)|^2\,dy,Wμ,d​(x)=Nμ,d​​Wμ,d​(x)​,Nμ,d​=∫R​∣Wμ,d​(y)∣2dy,

where, with qμ,r=π(2r+μ)q_{\mu,r}=\sqrt\pi(2r+\mu)qμ,r​=π​(2r+μ),

Wμ,d(y)=2π∑r∈Zexp⁡ ⁣(−(y−qμ,r)22d2−d2(y+qμ,r)28).W_{\mu,d}(y)=\sqrt{\frac2{\sqrt\pi}}\sum_{r\in\mathbb Z} \exp\!\left(-\frac{(y-q_{\mu,r})^2}{2d^2}-\frac{d^2(y+q_{\mu,r})^2}{8}\right).Wμ,d​(y)=π​2​​r∈Z∑​exp(−2d2(y−qμ,r​)2​−8d2(y+qμ,r​)2​).

The real square root in the denominator is viewed as a complex number. No condition on ddd or nonvanishing denominator is assumed, and this definition itself asserts no normalization property. Division by zero gives zero; infinite sums and Lebesgue integrals have value zero when their summability or integrability requirement fails.

QAlgorithms.Papers.Independent.Intakenext100.P65.numberMoment

For every complex sequence v:N→Cv:\mathbb N\to\mathbb Cv:N→C and every nonnegative integer kkk, this defines the real number

∑n=0∞nk∣vn∣2.\sum_{n=0}^{\infty}n^k|v_n|^2.n=0∑∞​nk∣vn​∣2.

There is no summability or normalization hypothesis; if the series is not summable, the total infinite-sum operation returns 000. When k=0k=0k=0, the expression is ∑n∣vn∣2\sum_n|v_n|^2∑n​∣vn​∣2, with 00=10^0=100=1; when k>0k>0k>0, its term at n=0n=0n=0 is zero.

QAlgorithms.Papers.Independent.Intakenext100.P65.permCode

For natural numbers d,Md,Md,M, an index 0≤μ<d0\le\mu<d0≤μ<d, and a configuration x∈{0,…,d−1}Mx\in\{0,\ldots,d-1\}^{M}x∈{0,…,d−1}M, this defines

permCode⁡d,M,μ(x)=∑m=0(d−1)Mexp⁡ ⁣(i 2πμmd)Cd,M(m)dM{1Cd,M(m),∑ixi=m,0,otherwise,\operatorname{permCode}_{d,M,\mu}(x) =\sum_{m=0}^{(d-1)M} \exp\!\left(i\,\frac{2\pi\mu m}{d}\right) \frac{\sqrt{C_{d,M}(m)}}{\sqrt{d^M}} \begin{cases} \displaystyle\frac1{\sqrt{C_{d,M}(m)}},&\sum_i x_i=m,\\ 0,&\text{otherwise}, \end{cases}permCoded,M,μ​(x)=m=0∑(d−1)M​exp(id2πμm​)dM​Cd,M​(m)​​⎩⎨⎧​Cd,M​(m)​1​,0,​∑i​xi​=m,otherwise,​

where Cd,M(m)C_{d,M}(m)Cd,M​(m) is the number of configurations y∈{0,…,d−1}My\in\{0,\ldots,d-1\}^{M}y∈{0,…,d−1}M with ∑iyi=m\sum_i y_i=m∑i​yi​=m. The upper summation bound uses natural subtraction, and inverses of zero are assigned zero, so any term with zero count contributes zero. An index μ\muμ requires d>0d>0d>0; M=0M=0M=0 is allowed and gives amplitude 111 on the unique empty configuration.

QAlgorithms.Papers.Independent.Intakenext100.P65.poissonWeight

For every real a,ga,ga,g and natural number lll, this defines

poissonWeight⁡(a,g,l)=(ga2)le−ga2l!.\operatorname{poissonWeight}(a,g,l) =\frac{(ga^2)^l e^{-ga^2}}{l!}.poissonWeight(a,g,l)=l!(ga2)le−ga2​.

Neither aaa nor ggg is restricted, so the definition does not require this quantity to be a probability. It includes l=0l=0l=0, when the value is e−ga2e^{-ga^2}e−ga2, including when ga2=0ga^2=0ga2=0.

QAlgorithms.Papers.Independent.Intakenext100.P65.qecBlock

For any pair of complex sequences B0,B1:N→CB_0,B_1:\mathbb N\to\mathbb CB0​,B1​:N→C, any real number ggg, and any nonnegative integers l,kl,kl,k, this defines a complex 2×22\times22×2 matrix whose entry at i,j∈{0,1}i,j\in\{0,1\}i,j∈{0,1} is

∑n=0∞[(n+ll) (g)l(1−g)n][(n+kk) (g)k(1−g)n](Bi)n+l‾ (Bj)n+k.\sum_{n=0}^{\infty} \left[\sqrt{\binom{n+l}{l}}\,(\sqrt g)^l(\sqrt{1-g})^n\right] \left[\sqrt{\binom{n+k}{k}}\,(\sqrt g)^k(\sqrt{1-g})^n\right] \overline{(B_i)_{n+l}}\,(B_j)_{n+k}.n=0∑∞​[(ln+l​)​(g​)l(1−g​)n][(kn+k​)​(g​)k(1−g​)n](Bi​)n+l​​(Bj​)n+k​.

No normalization, orthogonality, or summability hypotheses are imposed on B0,B1B_0,B_1B0​,B1​, and ggg is not restricted to [0,1][0,1][0,1]. Real square roots of negative arguments are zero, zeroth powers include 00=10^0=100=1, and each matrix entry is 000 if its displayed complex series is not summable.

QAlgorithms.Papers.Independent.Intakenext100.P65.qecC

For any two sequences of complex numbers B0,B1:N→CB_0,B_1:\mathbb N\to\mathbb CB0​,B1​:N→C, any real ggg, and any natural numbers l,kl,kl,k, including zero, define

al(n)=(n+ll) (g)l(1−g)n,sij=∑n=0∞al(n)ak(n) Bi(n+l)‾Bj(n+k).a_l(n)=\sqrt{\binom{n+l}{l}}\,(\sqrt g)^l(\sqrt{1-g})^n, \qquad s_{ij}=\sum_{n=0}^{\infty}a_l(n)a_k(n)\, \overline{B_i(n+l)}B_j(n+k).al​(n)=(ln+l​)​(g​)l(1−g​)n,sij​=n=0∑∞​al​(n)ak​(n)Bi​(n+l)​Bj​(n+k).

Then qecC(B,g,l,k)=Re⁡((s00+s11)/2)\mathrm{qecC}(B,g,l,k)=\operatorname{Re}((s_{00}+s_{11})/2)qecC(B,g,l,k)=Re((s00​+s11​)/2). No normalization, square summability, or convergence hypotheses are imposed on B0,B1B_0,B_1B0​,B1​, and ggg is not restricted to [0,1][0,1][0,1]. Each complex series is assigned zero if it is not summable; real square roots of negative numbers are zero.

QAlgorithms.Papers.Independent.Intakenext100.P65.qecX

For any two sequences of complex numbers B0,B1:N→CB_0,B_1:\mathbb N\to\mathbb CB0​,B1​:N→C, any real ggg, and any natural numbers l,kl,kl,k, including zero, define

al(n)=(n+ll) (g)l(1−g)n,sij=∑n=0∞al(n)ak(n) Bi(n+l)‾Bj(n+k).a_l(n)=\sqrt{\binom{n+l}{l}}\,(\sqrt g)^l(\sqrt{1-g})^n, \qquad s_{ij}=\sum_{n=0}^{\infty}a_l(n)a_k(n)\, \overline{B_i(n+l)}B_j(n+k).al​(n)=(ln+l​)​(g​)l(1−g​)n,sij​=n=0∑∞​al​(n)ak​(n)Bi​(n+l)​Bj​(n+k).

Then qecX(B,g,l,k)=Re⁡((s01+s10)/2)\mathrm{qecX}(B,g,l,k)=\operatorname{Re}((s_{01}+s_{10})/2)qecX(B,g,l,k)=Re((s01​+s10​)/2). No normalization, square summability, or convergence hypotheses are imposed on B0,B1B_0,B_1B0​,B1​, and ggg is not restricted to [0,1][0,1][0,1]. Each complex series is assigned zero if it is not summable; real square roots of negative numbers are zero.

QAlgorithms.Papers.Independent.Intakenext100.P65.qecZ

For any two sequences of complex numbers B0,B1:N→CB_0,B_1:\mathbb N\to\mathbb CB0​,B1​:N→C, any real ggg, and any natural numbers l,kl,kl,k, including zero, define

al(n)=(n+ll) (g)l(1−g)n,sij=∑n=0∞al(n)ak(n) Bi(n+l)‾Bj(n+k).a_l(n)=\sqrt{\binom{n+l}{l}}\,(\sqrt g)^l(\sqrt{1-g})^n, \qquad s_{ij}=\sum_{n=0}^{\infty}a_l(n)a_k(n)\, \overline{B_i(n+l)}B_j(n+k).al​(n)=(ln+l​)​(g​)l(1−g​)n,sij​=n=0∑∞​al​(n)ak​(n)Bi​(n+l)​Bj​(n+k).

Then qecZ(B,g,l,k)=Re⁡((s00−s11)/2)\mathrm{qecZ}(B,g,l,k)=\operatorname{Re}((s_{00}-s_{11})/2)qecZ(B,g,l,k)=Re((s00​−s11​)/2). No normalization, square summability, or convergence hypotheses are imposed on B0,B1B_0,B_1B0​,B1​, and ggg is not restricted to [0,1][0,1][0,1]. Each complex series is assigned zero if it is not summable; real square roots of negative numbers are zero.

QAlgorithms.Papers.Independent.Intakenext100.P65.qubitDickeCoordinates

For every natural numbers K,rK,rK,r and configuration x:{0,…,K−1}→{0,1}x:\{0,\ldots,K-1\}\to\{0,1\}x:{0,…,K−1}→{0,1}, this defines the complex amplitude

dK,r(x)={1(Kr),#{i:xi=1}=r,0,otherwise.d_{K,r}(x)= \begin{cases} \dfrac1{\sqrt{\binom K r}},& \#\{i:x_i=1\}=r,\\ 0,&\text{otherwise}. \end{cases}dK,r​(x)=⎩⎨⎧​(rK​)​1​,0,​#{i:xi​=1}=r,otherwise.​

Here the count is obtained by converting each coordinate to the Boolean assertion that its value is 111. No assumption r≤Kr\le Kr≤K is imposed: when r>Kr>Kr>K, every amplitude is zero. When K=0K=0K=0, the unique empty configuration has amplitude 111 if r=0r=0r=0 and 000 otherwise.

QAlgorithms.Papers.Independent.Intakenext100.P65.qubitFrobeniusSq

For any complex 2×22\times22×2 matrix AAA, the real number qubitFrobeniusSq(A)\mathrm{qubitFrobeniusSq}(A)qubitFrobeniusSq(A) is

qubitFrobeniusSq(A)=∑i=01∑j=01∣Aij∣2.\mathrm{qubitFrobeniusSq}(A)=\sum_{i=0}^1\sum_{j=0}^1|A_{ij}|^2.qubitFrobeniusSq(A)=i=0∑1​j=0∑1​∣Aij​∣2.

No additional assumptions are imposed on AAA.

QAlgorithms.Papers.Independent.Intakenext100.P65.quditFourierFactor

For arbitrary natural numbers d,μ,md,\mu,md,μ,m, with no bounds on μ\muμ or mmm, this defines

Fd(μ,m)=exp⁡ ⁣(i 2πμmd).F_d(\mu,m)=\exp\!\left(i\,\frac{2\pi\mu m}{d}\right).Fd​(μ,m)=exp(id2πμm​).

The division is real division before taking the complex exponential. The case d=0d=0d=0 is included: division by zero gives zero, so F0(μ,m)=1F_0(\mu,m)=1F0​(μ,m)=1.

QAlgorithms.Papers.Independent.Intakenext100.P65.regularizedLattice

For either μ∈{0,1}\mu\in\{0,1\}μ∈{0,1} and arbitrary real r,xr,xr,x, define

Vμ(r,x)=∑(a,b)∈Z2e−r(a2+b2)π−1/4e−(x−qa)2/2e iπ b(x−qa),qa=π(2a+μ).V_\mu(r,x) =\sum_{(a,b)\in\mathbb Z^2} e^{-r(a^2+b^2)}\pi^{-1/4} e^{-(x-q_a)^2/2} e^{\,i\sqrt\pi\,b(x-q_a)}, \qquad q_a=\sqrt\pi(2a+\mu).Vμ​(r,x)=(a,b)∈Z2∑​e−r(a2+b2)π−1/4e−(x−qa​)2/2eiπ​b(x−qa​),qa​=π​(2a+μ).

The real factors are embedded in C\mathbb CC. The sum is over integer pairs as a single family, with no assumption that r>0r>0r>0 or that this family is summable; its value is zero when the family is not summable.

QAlgorithms.Papers.Independent.Intakenext100.P65.residueCheck

For every natural number SSS and arbitrary complex sequence v:N→Cv:\mathbb N\to\mathbb Cv:N→C, this defines the complex sequence

(residueCheck⁡(S,v))n=exp⁡ ⁣(i 2πnS+1)vn(n∈N).(\operatorname{residueCheck}(S,v))_n =\exp\!\left(\mathrm i\,\frac{2\pi n}{S+1}\right)v_n \qquad(n\in\mathbb N).(residueCheck(S,v))n​=exp(iS+12πn​)vn​(n∈N).

No summability or normalization condition is imposed on vvv. The denominator is always positive, and the definition includes S=0S=0S=0.

QAlgorithms.Papers.Independent.Intakenext100.P65.residueLogicalZ

For every natural number SSS and arbitrary complex sequence v:N→Cv:\mathbb N\to\mathbb Cv:N→C, this defines

(residueLogicalZ⁡(S,v))n=exp⁡ ⁣(i πnS+1)vn(n∈N).(\operatorname{residueLogicalZ}(S,v))_n =\exp\!\left(\mathrm i\,\frac{\pi n}{S+1}\right)v_n \qquad(n\in\mathbb N).(residueLogicalZ(S,v))n​=exp(iS+1πn​)vn​(n∈N).

No summability or normalization assumption is required, and S=0S=0S=0 is included.

QAlgorithms.Papers.Independent.Intakenext100.P65.residueProject

For every natural number SSS, integer rrr, and complex sequence v:N→Cv:\mathbb N\to\mathbb Cv:N→C, this defines the complex sequence

(residueProject⁡(S,r,v))n={vn,n≡r(mod2(S+1)),0,otherwise.(\operatorname{residueProject}(S,r,v))_n= \begin{cases} v_n,&n\equiv r\pmod{2(S+1)},\\ 0,&\text{otherwise}. \end{cases}(residueProject(S,r,v))n​={vn​,0,​n≡r(mod2(S+1)),otherwise.​

Here nnn ranges over all nonnegative integers, while rrr may be negative; congruence is tested using integer remainders modulo the positive integer 2(S+1)2(S+1)2(S+1). No summability or normalization condition is imposed on vvv, and S=0S=0S=0 is included.

QAlgorithms.Papers.Independent.Intakenext100.P65.reverseInformation

For any function TTT from complex 2×22\times22×2 matrices to themselves, let EijE_{ij}Eij​ denote the matrix with entry 111 at (i,j)(i,j)(i,j) and zero elsewhere, and define JTJ_TJT​ by (JT)(o,i),(p,j)=T(Eij)op(J_T)_{(o,i),(p,j)}=T(E_{ij})_{op}(JT​)(o,i),(p,j)​=T(Eij​)op​, with all indices in {0,1}\{0,1\}{0,1}. Put σ=JT/2\sigma=J_T/2σ=JT​/2 and σout,op=∑i=01σ(o,i),(p,i)\sigma_{\mathrm{out},op}=\sum_{i=0}^1\sigma_{(o,i),(p,i)}σout,op​=∑i=01​σ(o,i),(p,i)​. Then

reverseInformation(T)=H(σout)−H(σ),H(M)=−Re⁡tr⁡ ⁣(f(M)),f(x)=xlog⁡2x,\mathrm{reverseInformation}(T)=H(\sigma_{\mathrm{out}})-H(\sigma), \qquad H(M)=-\operatorname{Re}\operatorname{tr}\!\bigl(f(M)\bigr),\quad f(x)=x\log_2x,reverseInformation(T)=H(σout​)−H(σ),H(M)=−Retr(f(M)),f(x)=xlog2​x,

where f(M)f(M)f(M) is formed using the continuous functional calculus. The definition requires no channel, linearity, positivity, or normalization assumptions on TTT.

QAlgorithms.Papers.Independent.Intakenext100.P65.smallBasis

For indices j,n∈{0,…,6}j,n\in\{0,\ldots,6\}j,n∈{0,…,6}, this defines the complex number smallBasis⁡(j,n)\operatorname{smallBasis}(j,n)smallBasis(j,n) as the nnn-th coordinate of bjb_jbj​, where, in coordinate order 0,…,60,\ldots,60,…,6,

b0=18(1,0,3,0,3,0,1),b1=18(1,0,−3,0,3,0,−1),b2=18(−3,0,−1,0,1,0,3),b3=18(−3,0,1,0,1,0,−3),b4=18(0,2,0,2,0,2,0),b5=18(0,−2,0,2,0,−2,0),b6=18(0,2,0,0,0,−2,0).\begin{aligned} b_0&=\tfrac1{\sqrt8}(1,0,\sqrt3,0,\sqrt3,0,1),\\ b_1&=\tfrac1{\sqrt8}(1,0,-\sqrt3,0,\sqrt3,0,-1),\\ b_2&=\tfrac1{\sqrt8}(-\sqrt3,0,-1,0,1,0,\sqrt3),\\ b_3&=\tfrac1{\sqrt8}(-\sqrt3,0,1,0,1,0,-\sqrt3),\\ b_4&=\tfrac1{\sqrt8}(0,\sqrt2,0,2,0,\sqrt2,0),\\ b_5&=\tfrac1{\sqrt8}(0,-\sqrt2,0,2,0,-\sqrt2,0),\\ b_6&=\tfrac1{\sqrt8}(0,2,0,0,0,-2,0). \end{aligned}b0​b1​b2​b3​b4​b5​b6​​=8​1​(1,0,3​,0,3​,0,1),=8​1​(1,0,−3​,0,3​,0,−1),=8​1​(−3​,0,−1,0,1,0,3​),=8​1​(−3​,0,1,0,1,0,−3​),=8​1​(0,2​,0,2,0,2​,0),=8​1​(0,−2​,0,2,0,−2​,0),=8​1​(0,2,0,0,0,−2,0).​

QAlgorithms.Papers.Independent.Intakenext100.P65.smallGenerator

This defines the complex 7×77\times77×7 matrix GGG, indexed by 0,…,60,\ldots,60,…,6, with diagonal

(G00,G11,G22,G33,G44,G55,G66)=(6,12,14,1,14,12,6),(G_{00},G_{11},G_{22},G_{33},G_{44},G_{55},G_{66}) =\left(6,\tfrac12,14,1,14,\tfrac12,6\right),(G00​,G11​,G22​,G33​,G44​,G55​,G66​)=(6,21​,14,1,14,21​,6),

off-diagonal entries

G04=G40=G26=G62=43,G15=G51=12,G_{04}=G_{40}=G_{26}=G_{62}=4\sqrt3, \qquad G_{15}=G_{51}=\tfrac12,G04​=G40​=G26​=G62​=43​,G15​=G51​=21​,

and all other entries zero. Equivalently, the definition embeds the tridiagonal matrices with neighboring entries (3/2,1,3/2)(\sqrt3/2,1,\sqrt3/2)(3​/2,1,3​/2) and (2/2,2/2)(\sqrt2/2,\sqrt2/2)(2​/2,2​/2) into the even and odd coordinates respectively, squares each embedded matrix, and adds eight times the first square to the second square.

QAlgorithms.Papers.Independent.Intakenext100.P65.smallUnitary

This defines

U=exp⁡ ⁣(2πi5G),U=\exp\!\left(\frac{2\pi i}{5}G\right),U=exp(52πi​G),

where the exponential is the matrix exponential and GGG is the complex matrix indexed by 0,…,60,\ldots,60,…,6 with diagonal (6,12,14,1,14,12,6)(6,\tfrac12,14,1,14,\tfrac12,6)(6,21​,14,1,14,21​,6), entries G04=G40=G26=G62=43G_{04}=G_{40}=G_{26}=G_{62}=4\sqrt3G04​=G40​=G26​=G62​=43​ and G15=G51=12G_{15}=G_{51}=\tfrac12G15​=G51​=21​, and every other off-diagonal entry zero. This declaration defines the matrix without itself asserting unitarity.

QAlgorithms.Papers.Independent.Intakenext100.P65.smearedKernelIntegral

For every μ,ν∈{0,1}\mu,\nu\in\{0,1\}μ,ν∈{0,1}, d∈Rd\in\mathbb Rd∈R, and z∈Cz\in\mathbb Cz∈C, define, for (r,s)∈Z2(r,s)\in\mathbb Z^2(r,s)∈Z2,

Kμν(r,s)=π2(2r+μ−ν+is),χμν(r,s)=exp⁡ ⁣(iπ(r+μ+ν2)s),w(a,b)=exp⁡ ⁣(ab‾−a‾b2).K_{\mu\nu}(r,s)=\sqrt{\frac\pi2}\bigl(2r+\mu-\nu+is\bigr),\qquad \chi_{\mu\nu}(r,s)=\exp\!\left(i\pi\left(r+\frac{\mu+\nu}{2}\right)s\right), \quad w(a,b)=\exp\!\left(\frac{a\overline b-\overline a b}{2}\right).Kμν​(r,s)=2π​​(2r+μ−ν+is),χμν​(r,s)=exp(iπ(r+2μ+ν​)s),w(a,b)=exp(2ab−ab​).

The declared complex quantity is

Iμν,d(z)=∑(r,s)∈Z2χμν(r,s)∫Cexp⁡ ⁣(−∣b∣2+∣Kμν(r,s)+b−z∣2d2)πd2/2 w(−b,z) w ⁣(z−b,Kμν(r,s)+b−z) db.I_{\mu\nu,d}(z)= \sum_{(r,s)\in\mathbb Z^2}\chi_{\mu\nu}(r,s) \int_{\mathbb C} \frac{\exp\!\left(-\frac{|b|^2+|K_{\mu\nu}(r,s)+b-z|^2}{d^2}\right)} {\pi d^2/2}\, w(-b,z)\,w\!\left(z-b,K_{\mu\nu}(r,s)+b-z\right)\,db.Iμν,d​(z)=(r,s)∈Z2∑​χμν​(r,s)∫C​πd2/2exp(−d2∣b∣2+∣Kμν​(r,s)+b−z∣2​)​w(−b,z)w(z−b,Kμν​(r,s)+b−z)db.

Here dbdbdb is Lebesgue measure on C≅R2\mathbb C\cong\mathbb R^2C≅R2, and the real exponential quotient is regarded as complex. No restriction on ddd, convergence, or integrability is assumed: division by zero gives zero, nonintegrable integrands have integral zero, and nonsummable outer series have sum zero.

QAlgorithms.Papers.Independent.Intakenext100.P65.smearedKernelSum

For every μ,ν∈{0,1}\mu,\nu\in\{0,1\}μ,ν∈{0,1}, d∈Rd\in\mathbb Rd∈R, and z∈Cz\in\mathbb Cz∈C, set

Kμν(r,s)=π2(2r+μ−ν+is),χμν(r,s)=exp⁡ ⁣(iπ(r+μ+ν2)s).K_{\mu\nu}(r,s)=\sqrt{\frac\pi2}\bigl(2r+\mu-\nu+is\bigr),\qquad \chi_{\mu\nu}(r,s)=\exp\!\left(i\pi\left(r+\frac{\mu+\nu}{2}\right)s\right).Kμν​(r,s)=2π​​(2r+μ−ν+is),χμν​(r,s)=exp(iπ(r+2μ+ν​)s).

Define

Sμν,d(z)=∑(r,s)∈Z2χμν(r,s)exp⁡ ⁣(−∣z−Kμν(r,s)∣22d2−d2∣z+Kμν(r,s)∣28).S_{\mu\nu,d}(z)= \sum_{(r,s)\in\mathbb Z^2}\chi_{\mu\nu}(r,s) \exp\!\left( -\frac{|z-K_{\mu\nu}(r,s)|^2}{2d^2} -\frac{d^2|z+K_{\mu\nu}(r,s)|^2}{8} \right).Sμν,d​(z)=(r,s)∈Z2∑​χμν​(r,s)exp(−2d2∣z−Kμν​(r,s)∣2​−8d2∣z+Kμν​(r,s)∣2​).

This definition includes d=0d=0d=0 and negative ddd, imposes no summability hypothesis, uses zero for division by zero, and assigns zero to a nonsummable series.

QAlgorithms.Papers.Independent.Intakenext100.P65.smearedNorm

For μ∈{0,1}\mu\in\{0,1\}μ∈{0,1} and every real ddd, put qμ,r=π(2r+μ)q_{\mu,r}=\sqrt\pi(2r+\mu)qμ,r​=π​(2r+μ) for r∈Zr\in\mathbb Zr∈Z and define

Wμ,d(x)=2π∑r∈Zexp⁡ ⁣(−(x−qμ,r)22d2−d2(x+qμ,r)28).W_{\mu,d}(x)=\sqrt{\frac2{\sqrt\pi}}\sum_{r\in\mathbb Z} \exp\!\left(-\frac{(x-q_{\mu,r})^2}{2d^2}-\frac{d^2(x+q_{\mu,r})^2}{8}\right).Wμ,d​(x)=π​2​​r∈Z∑​exp(−2d2(x−qμ,r​)2​−8d2(x+qμ,r​)2​).

The declared real quantity is

Nμ,d=∫R∣Wμ,d(x)∣2 dx.N_{\mu,d}=\int_{\mathbb R}|W_{\mu,d}(x)|^2\,dx.Nμ,d​=∫R​∣Wμ,d​(x)∣2dx.

The integral uses Lebesgue measure. There is no restriction on ddd, including d=0d=0d=0, and no positivity or integrability assertion is part of this definition; division by zero is assigned zero, a nonsummable series is assigned sum zero, and a nonintegrable function is assigned integral zero.

QAlgorithms.Papers.Independent.Intakenext100.P65.smearedWave

For either μ∈{0,1}\mu\in\{0,1\}μ∈{0,1} and arbitrary real d,xd,xd,x, define

Ψμ(d,x)=2π∑n∈Zexp⁡ ⁣(−(x−qn)22d2−d2(x+qn)28),qn=π(2n+μ),\Psi_\mu(d,x) =\sqrt{\frac{2}{\sqrt\pi}} \sum_{n\in\mathbb Z} \exp\!\left( -\frac{(x-q_n)^2}{2d^2} -\frac{d^2(x+q_n)^2}{8} \right), \qquad q_n=\sqrt\pi(2n+\mu),Ψμ​(d,x)=π​2​​n∈Z∑​exp(−2d2(x−qn​)2​−8d2(x+qn​)2​),qn​=π​(2n+μ),

with all real quantities embedded in C\mathbb CC. No positivity condition on ddd, normalization assertion, or summability hypothesis is included. Division by zero equals zero, and a nonsummable family has sum zero; consequently, at d=0d=0d=0 every summand is 111, so the defined value is zero.

QAlgorithms.Papers.Independent.Intakenext100.P65.spinX

For every natural number MMM, this defines a complex (M+1)×(M+1)(M+1)\times(M+1)(M+1)×(M+1) matrix indexed by i,j∈{0,…,M}i,j\in\{0,\ldots,M\}i,j∈{0,…,M}, with entries

(spinX⁡M)ij={(j+1)(M−j)2,i=j+1,(i+1)(M−i)2,j=i+1,0,otherwise,(\operatorname{spinX}_M)_{ij}= \begin{cases} \dfrac{\sqrt{(j+1)(M-j)}}2,&i=j+1,\\ \dfrac{\sqrt{(i+1)(M-i)}}2,&j=i+1,\\ 0,&\text{otherwise}, \end{cases}(spinXM​)ij​=⎩⎨⎧​2(j+1)(M−j)​​,2(i+1)(M−i)​​,0,​i=j+1,j=i+1,otherwise,​

where the displayed real values are regarded as complex numbers. In particular its diagonal is zero, and M=0M=0M=0 gives the zero 1×11\times11×1 matrix.

QAlgorithms.Papers.Independent.Intakenext100.P65.totalQuditExcitation

For arbitrary natural numbers d,Md,Md,M and a configuration x:{0,…,M−1}→{0,…,d−1}x:\{0,\ldots,M-1\}\to\{0,\ldots,d-1\}x:{0,…,M−1}→{0,…,d−1}, this defines its total excitation to be the natural number

∑i=0M−1xi.\sum_{i=0}^{M-1}x_i.i=0∑M−1​xi​.

For M=0M=0M=0, the unique empty configuration has total excitation zero, including when d=0d=0d=0; when d=0d=0d=0 and M>0M>0M>0, no such configuration exists.

QAlgorithms.Papers.Independent.Intakenext100.P65.traceReference

For any complex matrix RRR indexed by {0,1}×{0,1}\{0,1\}\times\{0,1\}{0,1}×{0,1}, traceReference(R)\mathrm{traceReference}(R)traceReference(R) is the complex 2×22\times22×2 matrix obtained by summing over the second coordinate:

traceReference(R)ij=R(i,0),(j,0)+R(i,1),(j,1),i,j∈{0,1}.\mathrm{traceReference}(R)_{ij} =R_{(i,0),(j,0)}+R_{(i,1),(j,1)}, \qquad i,j\in\{0,1\}.traceReference(R)ij​=R(i,0),(j,0)​+R(i,1),(j,1)​,i,j∈{0,1}.

No positivity or normalization assumption is imposed on RRR.

QAlgorithms.Papers.Independent.Intakenext100.P65.trajectoryScalar

For any real numbers k,gk,gk,g, natural number lll, and real times τ0,…,τl−1\tau_0,\ldots,\tau_{l-1}τ0​,…,τl−1​, this defines the real number

l! exp⁡ ⁣(−k2∑j=0l−1τj)(g)l.\frac{\sqrt{l!}\,\exp\!\left(-\frac{k}{2}\sum_{j=0}^{l-1}\tau_j\right)} {(\sqrt g)^l}.(g​)ll!​exp(−2k​∑j=0l−1​τj​)​.

There are no sign or ordering assumptions. The real square root is zero for g≤0g\le0g≤0, and division by zero is defined to give zero; consequently, the value is zero when l>0l>0l>0 and g≤0g\le0g≤0. When l=0l=0l=0, the empty sum is zero and the denominator is 111, so the value is 111 for every k,gk,gk,g.

QAlgorithms.Papers.Independent.Intakenext100.P65.weylPhase

For arbitrary complex numbers a,ba,ba,b, this defines

weylPhase⁡(a,b)=exp⁡ ⁣(ab‾−a‾b2).\operatorname{weylPhase}(a,b) =\exp\!\left(\frac{a\overline b-\overline a b}{2}\right).weylPhase(a,b)=exp(2ab−ab​).

QAlgorithms.Papers.Independent.Intakenext100.P65.worstPureFidelity

For any natural number ddd and any function TTT from complex d×dd\times dd×d matrices to themselves, define Pv=vv†P_v=vv^\daggerPv​=vv†, so (Pv)ij=vivj‾(P_v)_{ij}=v_i\overline{v_j}(Pv​)ij​=vi​vj​​. Then

worstPureFidelity(d,T)=inf⁡{Re⁡tr⁡ ⁣(PvT(Pv)):v∈Cd, ∥v∥2=1}.\mathrm{worstPureFidelity}(d,T) =\inf\left\{\operatorname{Re}\operatorname{tr}\!\bigl(P_vT(P_v)\bigr): v\in\mathbb C^d,\ \|v\|_2=1\right\}.worstPureFidelity(d,T)=inf{Retr(Pv​T(Pv​)):v∈Cd, ∥v∥2​=1}.

No channel or linearity assumption is required of TTT. The real infimum used here is zero when the set is empty or unbounded below; in particular, the value is zero for d=0d=0d=0.

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