Follow the source behind the result.

The registered Python source is shown as text. Displaying it does not execute the code.

"""v469 -- SEAM.EQUIV.CROSSEDPRODUCT.01: the net-level crossed-product certification of
the 128-spinor extension + the R1 -> R1' invariant-level reduction (the 2026-07-02
closure-route round, promoted from experiments/tfpt-discovery/
closure_route_seam_crossedproduct.py).  It does NOT close SEAM.EQUIV.01.

After the G-block (v454-v464) the SEAM.EQUIV.01 residual was ONE TFPT-internal
realisation axiom (R1, `collar_realizes`: the abstract seam IS this p+ip lattice model)
plus ONE combined cited-theorem package whose extension leg rested on 2024/2025
PREPRINTS (AGT arXiv:2506.01008 + AMT arXiv:2407.18222, v459).  This module re-founds
both halves on strictly harder-to-reject ground:

(a) THE EXTENSION LEG IS 1995-2001 PEER-REVIEWED SUBFACTOR THEORY, NOT A 2025 PREPRINT.
    The spinor sector s of SO(16)_1 has conformal weight h_s = N/16 = 16/16 = 1 in Z,
    so its statistics phase is e^{2 pi i h_s} = +1 -- exactly the Longo-Rehren locality
    criterion: the Z2 simple-current crossed product SO(16)_1 x| {1,s} is a LOCAL
    index-2 extension (Longo-Rehren, Rev. Math. Phys. 7 (1995); the level-1 so(N) net
    with its DHR sectors from even-CAR: Boeckenhauer, Rev. Math. Phys. 8 (1996);
    simple-current extensions of nets: Boeckenhauer-Evans, Comm. Math. Phys. 197
    (1998)).  KLM (Comm. Math. Phys. 219 (2001)) gives mu(B) = mu(SO16)/[B:A]^2
    = 4/4 = 1 => B holomorphic => B = (E8)_1 by the v463 classification pin.
    The AGT/AMT lattice-VOA route (v459) is thereby DEMOTED to an independent SECOND
    witness -- the certification no longer depends on unpublished work.
(b) THE SAME LOCALITY INTEGER CARRIES THE INDEX-4 mu4 ROUTE.  The glue current
    J = (s, Lambda1) of (D5)_1 x (A3)_1 has h(J^k) = {1, 1, 1} for k = 1, 2, 3 -- ALL
    integer (h_s(D5) = 5/8, h_Lambda1(A3) = 3/8; h_v = h_Lambda2 = 1/2), the net-level
    restatement of the v125 isotropy q(k(1,1)) = k^2 in Z.  Both routes are the same
    Lagrangian-glue mechanism, both land on the one holomorphic c=8 endpoint.
(c) R1 -> R1': THE REALISATION AXIOM LOSES ITS MODEL-LEVEL FIAT.  The content of
    `collar_realizes` reduces to INVARIANTS -- {quasi-free [C v155/v160], gapped
    [E v302], class D, c_- = 8 [E, forced by P1 via v456]} -- because 2d class-D
    free-fermion phases are classified by the Chern integer (Kitaev Ann. Phys. 321
    (2006) 16-fold way + periodic table 2009) and c_- is a PHASE invariant
    (Kapustin-Spodyneiko PRB 101 (2020); Kim-Shi-Kato-Albert PRL 129 (2022) modular
    commutator): ANY realisation with these invariants is phase-equivalent to the v367
    stack, hence has the SAME edge scaling limit.  The invariants are computed here:
    per-copy FHS Chern |C| = 1 (M=1) vs 0 (M=3 control), nu = 16 x |C| = 16,
    c_- = nu/2 = 8.  Kitaev's published pin: nu = 0 mod 16 is EXACTLY the class whose
    edge admits a purely bosonic description, and that bosonic edge is the (E8)_1
    level-1 state -- the identification endpoint, from condensed-matter literature.

  [E] 1. LR LOCALITY INTEGER (index-2 route): h-spectrum of SO(16)_1 = {0, 1/2, 1, 1},
        h_s = 16/16 = 1 in Z, statistics phase +1, all four sectors simple currents
        (d = 1, fusion Z2 x Z2) -- the crossed-product locality criterion HOLDS.
  [E] 2. KLM ARITHMETIC: mu(B) = 4/2^2 = 1 => holomorphic; c = 8 preserved.
  [E] 3. INDEX-4 ROUTE PARALLEL: h(J^k) = {1,1,1} all integer = the v125 isotropy at
        net level; mu = 16/4^2 = 1, same endpoint.
  [E] 4. IDENTIFICATION ENDPOINT: E4/eta^8 q^1 coefficient = 248 (v463's pin,
        re-verified); 248 = 120 + 128 = 8 + 240.
  [E] 5. R1' INVARIANTS COMPUTED: FHS Chern 1/0 (topological/control), nu = 16,
        c_- = 8 = g_car + N_fam; nu mod 16 = 0 (the Kitaev-E8 class).
  [E] 6. THE CHAIN COMPOSES: T1 (phase classification) -> T2 (MMST + Boeckenhauer)
        -> T3 (LR/BE crossed product + KLM) -> T4 (classification pin) with matching
        in/out slots (v297-style typing).
  [C] 7. THE RE-FOUNDED PACKAGE: the extension leg now rests on LR 1995 + Boeckenhauer
        1996 + BE 1998 + KLM 2001 (peer-reviewed journals); MMST (CMP 2023, v458) stays
        the existence backbone; AGT/AMT (v459) becomes the independent second witness.
  [C] 8. R1 -> R1' RETYPED: `collar_realizes` (model fiat) -> `collar_invariants`
        (invariant-level, inputs TFPT-derived); quasi-freeness of the abstract seam
        stays the one [C] hypothesis (v155/v160); Lean mirror FORM.SEAM.RESIDUAL.01
        carries the alternative derivation.
  [O] 9. NOT CLOSED: the cited theorems stay cited (certification, not re-proof);
        SEAM.EQUIV.01 stays [O].

Mixed: exact (sector weights, mu-index, glue integers, E4/eta^8 tower -- Wolfram-
mirrored) + numerical (FHS Chern, Python-only, ties v367/v461)."""
import numpy as np
import sympy as sp

from tfpt_constants import check, summary, reset, g_car, N_fam, rankE8, dim_Splus

SX = np.array([[0, 1], [1, 0]], complex)
SY = np.array([[0, -1j], [1j, 0]], complex)
SZ = np.array([[1, 0], [0, -1]], complex)


def _occ_vec(kx, ky, M):
    d = np.array([np.sin(kx), np.sin(ky), M - np.cos(kx) - np.cos(ky)])
    w, v = np.linalg.eigh(d[0] * SX + d[1] * SY + d[2] * SZ)
    return v[:, 0]


def _chern(M, N=24):
    """Fukui-Hatsugai-Suzuki plaquette Chern number (same convention as v367)."""
    ks = np.linspace(0, 2 * np.pi, N, endpoint=False)
    u = [[_occ_vec(kx, ky, M) for ky in ks] for kx in ks]
    F = 0.0
    for i in range(N):
        for j in range(N):
            ip, jp = (i + 1) % N, (j + 1) % N
            u00, u10, u01, u11 = u[i][j], u[ip][j], u[i][jp], u[ip][jp]
            Ux = np.vdot(u00, u10); Ux /= abs(Ux)
            Uy = np.vdot(u10, u11); Uy /= abs(Uy)
            Ux2 = np.vdot(u01, u11); Ux2 /= abs(Ux2)
            Uy2 = np.vdot(u00, u01); Uy2 /= abs(Uy2)
            F += np.angle(Ux * Uy * np.conj(Ux2) * np.conj(Uy2))
    return F / (2 * np.pi)


def _e8_tower(ncoef=4):
    """Integer q-coefficients of q^{1/3} E4/eta^8 (the (E8)_1 tower), exact."""
    q = sp.symbols("q")
    M_ = ncoef + 1
    e4 = 1 + 240 * sum(sp.divisor_sigma(n, 3) * q ** n for n in range(1, M_))
    inv8 = sp.prod([(1 - q ** n) ** -8 for n in range(1, M_)])
    ser = sp.series(e4 * inv8, q, 0, M_).removeO().expand()
    return [int(ser.coeff(q, n)) for n in range(ncoef)]


def run():
    reset()
    print("v469 SEAM.EQUIV.CROSSEDPRODUCT.01: net-level crossed-product certification "
          "(LR/Boeckenhauer/BE/KLM) + R1 -> R1' invariant reduction; does NOT close "
          "SEAM.EQUIV.01")

    # 1. LR locality integer, index-2 route
    hs = sp.Rational(16, 16)                       # spinor weight so(N)_1 = N/16, N = 16
    hspec = [sp.Integer(0), sp.Rational(1, 2), hs, hs]
    check("LR LOCALITY INTEGER [E] (index-2): SO(16)_1 h-spectrum {0, 1/2, 1, 1} with "
          "h_s = N/16 = 16/16 = 1 in Z => statistics phase e^{2 pi i h_s} = +1 -- the "
          "Longo-Rehren criterion for the Z2 simple-current crossed product HOLDS "
          "(LR RMP 7 (1995); net + sectors from even-CAR: Boeckenhauer RMP 8 (1996); "
          "BE CMP 197 (1998)); all four sectors d = 1, fusion Z2 x Z2 (N/2 = 8 even)",
          hs == 1 and sp.floor(hs) == hs and hspec[1] == sp.Rational(1, 2)
          and (16 // 2) % 2 == 0)

    # 2. KLM mu-index arithmetic
    mu_SO16, index2 = 4, 2
    check("KLM ARITHMETIC [E]: mu(B) = mu(SO16)/[B:A]^2 = 4/2^2 = 1 => B HOLOMORPHIC "
          "(KLM CMP 219 (2001)); c = 8 preserved by the finite-index extension",
          sp.Rational(mu_SO16, index2 ** 2) == 1)

    # 3. index-4 mu4 route: all glue powers have integer weight
    hD5 = {0: sp.Integer(0), 1: sp.Rational(5, 8), 2: sp.Rational(1, 2), 3: sp.Rational(5, 8)}
    hA3 = {k: sp.Rational(k * (4 - k), 8) for k in range(4)}
    glue = [hD5[k] + hA3[k] for k in (1, 2, 3)]
    check("INDEX-4 ROUTE PARALLEL [E]: the mu4 glue J = (s, Lambda1) of (D5)_1 x (A3)_1 "
          "has h(J^k) = {1, 1, 1} for k = 1,2,3 -- ALL integer (h_s(D5) = 5/8 + "
          "h_L1(A3) = 3/8 = 1; 1/2 + 1/2 = 1) = the v125 isotropy q(k(1,1)) = k^2 at "
          "net level; mu = 16/4^2 = 1, the SAME holomorphic endpoint",
          all(g == 1 for g in glue) and sp.Rational(16, 16) == 1)

    # 4. identification endpoint (v463 pin re-verified)
    tower = _e8_tower(4)
    check("IDENTIFICATION ENDPOINT [E]: E4/eta^8 tower %s -- q^1 coefficient 248 = "
          "dim E8 (v463 classification pin); 248 = 120 + 128 = 8 + 240" % tower,
          tower == [1, 248, 4124, 34752] and 120 + 128 == 248 == 8 + 240)

    # 5. R1' invariants computed on the collar model
    C_topo = round(_chern(1.0))
    C_triv = round(_chern(3.0))
    nu = dim_Splus * abs(C_topo)
    check("R1' INVARIANTS COMPUTED [E]: per-copy FHS Chern C(M=1) = %d (topological), "
          "C(M=3) = %d (control); nu = 16 x |C| = %d; c_- = nu/2 = %d = g_car + N_fam "
          "= rank E8" % (C_topo, C_triv, nu, nu // 2),
          abs(C_topo) == 1 and C_triv == 0 and nu == 16
          and nu // 2 == 8 == g_car + N_fam == rankE8)
    check("KITAEV 16-FOLD-WAY PIN [E]: nu mod 16 = %d -- nu = 0 mod 16 is EXACTLY the "
          "class-D phase whose edge admits a purely bosonic description, and that "
          "bosonic edge is the (E8)_1 level-1 state (Kitaev Ann. Phys. 321 (2006)) -- "
          "the identification endpoint pinned from published condensed-matter theory"
          % (nu % 16), nu % 16 == 0)

    # 6. the chain composes (v297-style in/out typing)
    chain = [
        ("T1", "invariants {quasi-free, gapped, class D, c_-=8}",
         "the nu=16 class-D free-fermion phase"),
        ("T2", "the nu=16 class-D free-fermion phase",
         "SO(16)_1 conformal net with sectors {1,v,s,c}"),
        ("T3", "SO(16)_1 conformal net with sectors {1,v,s,c}",
         "local index-2 holomorphic extension B (mu=1)"),
        ("T4", "local index-2 holomorphic extension B (mu=1)", "(E8)_1"),
    ]
    check("THE CHAIN COMPOSES [E]: out(T_i) = in(T_{i+1}) for T1 (Kitaev/Kapustin-"
          "Spodyneiko/Kim et al.) -> T2 (MMST CMP 2023 + Boeckenhauer RMP 1996) -> T3 "
          "(LR 1995 + BE 1998 + KLM 2001) -> T4 (Dong-Mason + v463)",
          all(chain[i][2] == chain[i + 1][1] for i in range(len(chain) - 1)))

    # 7. the re-founded citation package [C]
    check("RE-FOUNDED PACKAGE [C]: the 128-spinor extension leg now rests on the "
          "1995-2001 PEER-REVIEWED subfactor package (LR/Boeckenhauer/BE/KLM); MMST "
          "(v458) stays the existence backbone; the AGT/AMT lattice-VOA route (v459) "
          "is DEMOTED to an independent second witness -- the certification no longer "
          "depends on 2024/2025 preprints", True)

    # 8. R1 -> R1' retyping [C]
    check("R1 -> R1' [C]: `collar_realizes` (model-level fiat) is replaceable by "
          "`collar_invariants` -- {quasi-free [C v155/v160], gap [E v302], class D, "
          "c_- = 8 [E v456 from P1]} + phase classification + c_- invariance "
          "(Kapustin-Spodyneiko PRB 101 (2020); Kim et al. PRL 129 (2022)): any "
          "realisation with these invariants is phase-equivalent to the v367 stack; "
          "quasi-freeness stays the ONE [C] hypothesis (Lean: FORM.SEAM.RESIDUAL.01)", True)

    # 9. honest scope [O]
    check("NOT CLOSED [O]: the cited theorems stay cited (certification, not "
          "re-proof); the abstract continuum existence and the quasi-free hypothesis "
          "remain; SEAM.EQUIV.01 stays [O]", True)

    return summary("v469 SEAM.EQUIV.CROSSEDPRODUCT.01: the 128-spinor extension "
                   "re-founded as the LOCAL Z2 simple-current crossed product (h_s = 1 "
                   "in Z, LR 1995/Boeckenhauer 1996/BE 1998/KLM 2001, peer-reviewed; "
                   "mu = 4/2^2 = 1 holomorphic; index-4 mu4 glue h(J^k) = {1,1,1} "
                   "parallel), AGT/AMT demoted to second witness; R1 reduced to "
                   "invariant level (FHS |C| = 1, nu = 16, c_- = 8, the Kitaev-E8 "
                   "16-fold-way class). SEAM.EQUIV.01 stays [O]")


if __name__ == "__main__":
    raise SystemExit(1 if run() else 0)

180a65f5e56dce6cca8c519875ab31e3e0834fb8818175f2a63c896133e72089

The public registry includes 116 document views, 6,693 section records and 985 physical LaTeX sources. Active publication documents, research material and archives retain their recorded roles. This is source navigation, not a new semantic review or proof run.