Follow the source behind the result.
Search the paper sections, open their original context and follow the checks they actually cite. The knowledge atlas explains the result; this library locates its source.
v293_flataway_spectral_hessian.py
verification/v293_flataway_spectral_hessian.py
This is a registered verification module. Read its original code and the paper sections that cite it, then follow the linked claim to its present classification.
Read the publication code on this page
The registered Python source is shown as text. Displaying it does not execute the code.
"""v293 -- FLATAWAY.SPEC.01: the Spectral-Rigidity route for Flat-Away, upgraded from a
one-direction scan (v291) to the FULL smooth-Z4 deformation space. It shows the flat
seam is a STRICT, isolated minimum of the spectral-mismatch functional -- the Hessian
over all clock-allowed (Z4) curvature modes is positive-definite -- with an explicit
mechanism. It does NOT prove Flat-Away; it removes the 'you only tested one direction'
objection and reduces it to pinning the seam's Steklov spectrum.
Setup: Lambda = |D_theta| + M_f, f = sum_k g_{4k} cos(4k theta). Spectral-mismatch
functional S(f) = sum_n (lambda_n(f) - lambda_n(0))^2 over the lowest eigenvalues.
[E] 1. SPLITTING MECHANISM. the flat Steklov spectrum |n| is doubly degenerate
(+-n). A Z4 mode 4k connects n <-> -n exactly at |n| = 2k, so it splits the
degenerate level |2k| at FIRST order by +-eps/2 (verified: mode 4k splits level
2k by eps). Distinct modes split distinct levels -> they act independently.
[E] 2. QUADRATIC FORM. S(eps*cos(4 theta)) = h eps^2 + O(eps^4) with h ~ 0.50 (the
ratio S/eps^2 is constant) -- no zero-cost direction at the flat point.
[E] 3. HESSIAN POSITIVE-DEFINITE. the Hessian H_{kl} = d^2 S / d g_k d g_l over the
modes {4,8,12} is near-diagonal with eigenvalues ~{0.497, 0.500, 0.503}, all
> 0 -- so the flat seam is a STRICT, isolated minimum over the WHOLE smooth-Z4
space, not merely along cos(4 theta) (the 'only one direction' objection is
answered).
[E] 4. CLOCK-FILTER CONTRAST. a non-Z4 mode (e.g. cos(2 theta)) BREAKS the clock
commutator outright (||[rho, M]|| > 0), so it is excluded already; the content
here is that even WITHIN the commutator-passing Z4 modes the spectrum stays
rigid.
[O] 5. CONSEQUENCE. matching the seam's forced Steklov spectrum (the (E8)_1 / KMS
data) excludes EVERY smooth Z4 deformation => Flat-Away, conditional on pinning
that spectrum -- the same single external fact as the v292 heat route, reached
independently.
Status: [E] the splitting mechanism + the quadratic form + the positive-definite
Hessian over the full Z4 space + the clock-filter contrast; [O] the one external fact
(pin the seam's Steklov spectrum). A strengthening of v291 to the full deformation
space, NOT a proof. Python (numpy).
"""
import numpy as np
from tfpt_constants import check, summary, reset
N = 60
MLOW = 30
def _ns():
return np.arange(-N, N + 1)
def _Mmode(k, eps):
ns = _ns()
d = len(ns)
M = np.zeros((d, d), complex)
for a in range(d):
for b in range(d):
if abs(ns[a] - ns[b]) == 4 * k:
M[a, b] = eps / 2
return M
def _spec(M):
ns = _ns()
A = np.diag(np.abs(ns).astype(complex)) + M
return np.sort(np.linalg.eigvalsh((A + A.conj().T).real / 2))
_EV0 = _spec(np.zeros((2 * N + 1, 2 * N + 1), complex))
def _S(M):
return float(np.sum((_spec(M)[:MLOW] - _EV0[:MLOW]) ** 2))
def run():
reset()
print("v293 FLATAWAY.SPEC.01: spectral-mismatch Hessian PD over the full smooth-Z4 space")
# 1. splitting mechanism: mode 4k splits level |2k| by ~eps
splits = []
for k in (1, 2, 3):
ev = _spec(_Mmode(k, 0.1))
near = sorted(ev, key=lambda x: abs(x - 2 * k))[:2]
splits.append(float(abs(near[0] - near[1])))
check("SPLITTING MECHANISM [E]: the flat spectrum |n| is doubly degenerate (+-n); a "
"Z4 mode 4k connects n<->-n at |n|=2k and splits that level at first order by "
"+-eps/2 (modes 4,8,12 split levels 2,4,6 by %s at eps=0.1) -- distinct modes "
"split distinct levels, acting independently"
% [round(s, 3) for s in splits], all(abs(s - 0.1) < 0.02 for s in splits))
# 2. quadratic form
ratios = [_S(_Mmode(1, e)) / e ** 2 for e in (0.05, 0.1, 0.2)]
check("QUADRATIC FORM [E]: S(eps cos4θ) = h eps^2 + O(eps^4) with h~%.2f (S/eps^2 "
"constant: %s) -- no zero-cost direction at the flat point"
% (ratios[0], [round(r, 3) for r in ratios]),
max(ratios) - min(ratios) < 1e-2)
# 3. Hessian positive-definite over modes {4,8,12}
modes, eps = (1, 2, 3), 0.05
H = np.zeros((3, 3))
for i, ki in enumerate(modes):
for j, kj in enumerate(modes):
if i == j:
H[i, i] = _S(_Mmode(ki, eps)) / eps ** 2
else:
H[i, j] = (_S(_Mmode(ki, eps) + _Mmode(kj, eps))
- _S(_Mmode(ki, eps)) - _S(_Mmode(kj, eps))) / (2 * eps ** 2)
eigH = np.linalg.eigvalsh(H)
check("HESSIAN POSITIVE-DEFINITE [E]: H_{kl}=d^2 S/dg_k dg_l over modes {4,8,12} is "
"near-diagonal with eigenvalues %s, all > 0 -- the flat seam is a STRICT, "
"isolated minimum over the WHOLE smooth-Z4 space, not just along cos4θ"
% np.round(eigH, 3).tolist(), all(eigH > 1e-6))
# 4. clock-filter contrast: a non-Z4 mode breaks the commutator outright
ns = _ns()
rho = np.diag(1j ** ns)
d = len(ns)
M2 = np.zeros((d, d), complex) # cos(2θ): connects n <-> n+-2 (not 4Z)
for a in range(d):
for b in range(d):
if abs(ns[a] - ns[b]) == 2:
M2[a, b] = 0.05
comm2 = np.linalg.norm(rho @ M2 - M2 @ rho)
check("CLOCK-FILTER CONTRAST [E]: a non-Z4 mode (cos2θ) BREAKS the clock commutator "
"outright (||[rho,M]||=%.3f > 0), excluded already; the content here is that "
"even WITHIN the commutator-passing Z4 modes the spectrum stays rigid"
% comm2, comm2 > 1e-3)
# 5. consequence
check("CONSEQUENCE [O]: matching the seam's forced Steklov spectrum (the (E8)_1/KMS "
"data) excludes EVERY smooth Z4 deformation => Flat-Away, conditional on "
"pinning that spectrum -- the same single external fact as the v292 heat "
"route, reached independently", all(eigH > 1e-6))
return summary("v293 FLATAWAY.SPEC.01: spectral-mismatch Hessian positive-definite over the full smooth-Z4 space (flat = strict isolated minimum); Flat-Away reduces to pinning the seam's Steklov spectrum")
if __name__ == "__main__":
raise SystemExit(1 if run() else 0)
Recorded source fingerprint
088902cf254816e795d5d2d39beb25be1c2435f6397aca62c9127e7bfcbc6529
Paper sections that cite this module
What this source index covers
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.