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.
v299_hypergraph_0d_autonomous.py
verification/v299_hypergraph_0d_autonomous.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.
"""v299 -- the hypergraph / Wolfram-program bridge, hardened. This does NOT claim
"P2 is derived" or "E8 emerges from nothing". It proves the precise, honest statement:
P2 (the 3-arm (2,3,r) exceptional branch type) is the irreducible seed on which a
LOCAL, witness-based spectral growth dynamics REACHES E8 = T(2,3,5) as the LAST FINITE
point: it runs E6 -> E7 -> E8 and stops at the affine boundary Ê8 (the critical rho=2
edge). The Wolfram floor and the TFPT floor fall back to the SAME (2,3,5) seed.
On the groups (a deliberate precision): the icosahedron GRAPH has automorphism group
I_h = A5 x Z2 of order 120 (rotations A5 order 60 + reflections). The McKay group is the
binary icosahedral group 2I, the SU(2) double cover of the rotation group A5 -- also order
120 but a DIFFERENT (perfect) group; it is 2I, not the graph automorphism group, whose
McKay graph is affine E8 (built from quaternions in v219).
Negative controls (the built-in guards against over-reach): a path A_n seed and a D_n seed
have the EXACT closed forms rho(A_n)=2cos(pi/(n+1)) and rho(D_n)=2cos(pi/(2n-2)), both < 2
for ALL n -- so rho<=2 selects the ADE world, NOT E8; only the exceptional (1,2,r) arm does.
Claims (hard-typed):
[E] HYP.ICO.01 icosahedron = 3-uniform hypergraph (12 V, 20 triangular hyperedges,
30 edges = h(E8)); |Aut(graph)| = 120 = |I_h| = |A5 x Z2| (the McKay
group is the SU(2) binary lift 2I, a different order-120 group; v219).
[E] HYP.REWRITE.01 the 1->4 subdivision rewrite preserves all 120 icosahedral
symmetries at every scale (V: 12 -> 42 -> 162).
[E] HYP.FUSION.01 the 2I fusion tensor N^k_ij is a nonneg-integer weighted 3-uniform
hypergraph on 9 nodes; its 2-dim-rep slice = the affine E8 McKay
graph (A.deg = 2 deg, top eigenvalue 2).
[E] HYP.SMITH.01 Smith: rho<=2 selects (affine) ADE; rho(A_n,E6,E7,E8) < 2,
rho(Ê8) = 2, rho(E10) > 2; E8 = maximal finite exceptional in
T(2,3,r).
[E] HYP.CW.01 Collatz-Wielandt: rho<=2 <=> a positive witness m with A m <= 2m
exists (a LOCAL per-node balance), holds for E8/Ê8, fails for E10.
[E*] HYP.AUTO.01 the autonomous rule (self-generated witness via projectively-
normalised diffusion + LOCAL gate A m <= 2m, NO global rho) runs
E6 -> E7 -> E8 -> Ê8 then stops -- exact UNDER the 3-arm (2,3,*)
seed and the grow-longest-leg policy. Critical witness = the Kac
marks.
[E] HYP.NEG.AN.01 control: rho(A_n)=2cos(pi/(n+1))<2 for ALL n (closed form), so the
A-family never reaches E8.
[E] HYP.NEG.DN.01 control: rho(D_n)=2cos(pi/(2n-2))<2 for ALL n (closed form), so the
D-family never reaches E8; only the (1,2,r) arm does.
[O] HYP.SEED.01 the 3-arm (2,3,r) exceptional branch is the SINGLE irreducible input
= TFPT's P2 / the icosahedral choice (NOT "P2 derived").
Status: [E] the hypergraph facts (ICO/REWRITE/FUSION), the spectral selection
(SMITH/CW), the two negative controls; [E*] the autonomous rule under the stated seed +
policy; [O] the seed itself = P2. A second, network-theoretic reading of the SAME
(2,3,5)/E8 core, honestly bounded. Builds on v219/v298. Python (numpy; imports v219
for the 2I group + characters).
"""
import itertools
import numpy as np
import v219_icosahedral_mckay as mck
from tfpt_constants import check, summary, reset, N_fam, g_car
PHI = (1 + np.sqrt(5)) / 2
# ---------- icosahedron (3-uniform hypergraph) ----------------------------
def _icosahedron():
V = []
for a, b in itertools.product([1, -1], repeat=2):
V += [(0, a, b * PHI), (a, b * PHI, 0), (b * PHI, 0, a)]
V = np.array(sorted(set(map(tuple, np.round(V, 9)))))
D = np.round(((V[:, None] - V[None]) ** 2).sum(-1), 6)
el2 = np.min(D[D > 1e-9])
A = (np.abs(D - el2) < 1e-6).astype(int)
faces = [t for t in itertools.combinations(range(len(V)), 3)
if A[t[0], t[1]] and A[t[0], t[2]] and A[t[1], t[2]]]
return V, A, faces
def _graph_autos(A):
n = len(A)
nbr = [set(np.where(A[i])[0]) for i in range(n)]
autos = []
def bt(perm, used):
k = len(perm)
if k == n:
autos.append(tuple(perm))
return
for img in range(n):
if img in used:
continue
if all((perm[j] in nbr[img]) == (j in nbr[k]) for j in range(k)):
bt(perm + [img], used | {img})
bt([], set())
return autos
# ---------- star graphs T_{p,q,r} (Dynkin) -------------------------------
def _star(legs):
nodes = 1
adj = []
for L in legs:
prev = 0
for _ in range(L):
adj.append((prev, nodes))
prev = nodes
nodes += 1
A = np.zeros((nodes, nodes))
for a, b in adj:
A[a, b] = A[b, a] = 1
return A
def _rho(A):
return float(max(np.linalg.eigvalsh(A)))
def _witness_localgate(A):
"""Self-generate the Perron witness by projectively-normalised lazy diffusion, then
test the LOCAL per-node balance A m <= 2 m (no global rho in the decision)."""
d = A.shape[0]
M = (A + 2 * np.eye(d)) / 4
rng = np.random.default_rng(0)
m = rng.random(d)
for _ in range(5000):
m = M @ m
m = m / m.min() # projective normalisation
return m, bool(np.all(A @ m <= 2 * m + 1e-6))
def _grow(seed, policy_longest=True, cap=12):
"""Grow the longest leg while the LOCAL gate accepts; return the leg-sequence."""
legs = list(seed)
seq = []
for _ in range(cap):
seq.append(tuple(sorted(legs)))
legs2 = legs[:]
legs2[legs2.index(max(legs2))] += 1
if _witness_localgate(_star(legs2))[1]:
legs = legs2
else:
break
return seq
def run():
reset()
print("v299 hypergraph/Wolfram bridge: the (2,3,r) seed reaches E8 as the last finite point of a local ADE rule")
# HYP.ICO.01
V, A, faces = _icosahedron()
nV, nE, nF = len(V), int(A.sum() // 2), len(faces)
naut = len(_graph_autos(A))
check("HYP.ICO.01 [E]: the icosahedron is a 3-uniform hypergraph (V=%d, triangular "
"hyperedges=%d, edges=%d=h(E8)); |Aut(graph)|=%d=|I_h|=|A5xZ2| (rotations A5 "
"order 60 + reflections). The McKay group is the SU(2) binary lift 2I (order "
"120, perfect, != I_h), built from quaternions in v219" % (nV, nF, nE, naut),
nV == 12 and nF == 20 and nE == 30 == 2 * N_fam * g_car and naut == 120)
# HYP.REWRITE.01: 1->4 subdivision preserves all 120 symmetries
autos = _graph_autos(A)
Os = [np.linalg.lstsq(V, V[list(s)], rcond=None)[0].T for s in autos]
nbr = [set(np.where(A[i])[0]) for i in range(12)]
edges = sorted({(min(i, j), max(i, j)) for i in range(12) for j in nbr[i]})
scale = np.linalg.norm(V[0])
mids = [((V[i] + V[j]) / 2) / np.linalg.norm((V[i] + V[j]) / 2) * scale for i, j in edges]
Vsub = np.vstack([V, np.array(mids)])
preserved = sum(all(np.min(np.abs(Vsub - w).sum(1)) < 1e-6 for w in (O @ Vsub.T).T)
for O in Os)
check("HYP.REWRITE.01 [E]: the 1->4 triangle-subdivision rewrite is equivariant -- "
"all %d/%d full icosahedral symmetries (I_h, whose rotation subgroup A5 has "
"the binary lift 2I) still permute the subdivided hypergraph (V 12->%d->162); "
"the icosahedral symmetry is carried to every scale"
% (preserved, len(Os), len(Vsub)), preserved == 120 and len(Vsub) == 42)
# HYP.FUSION.01: 2I fusion tensor (via v219) is a 3-uniform hypergraph; V-slice = E8
G = mck._icosians()
n = len(G)
index = {mck._key(q): i for i, q in enumerate(G)}
table = np.empty((n, n), int)
for i in range(n):
for j in range(n):
table[i, j] = index[mck._key(mck._quat_mult(G[i], G[j]))]
classes = mck._conjugacy_classes(table, n)
degs, chars, _ = mck._dixon(table, classes, n)
chars = np.array(chars)
sizes = np.array([len(c) for c in classes])
nc = len(classes)
Nt = np.zeros((nc, nc, nc))
for i in range(nc):
for j in range(nc):
for k in range(nc):
Nt[i, j, k] = np.real(np.sum(sizes * chars[i] * chars[j] * np.conj(chars[k])) / n)
Nt = np.round(Nt).astype(int)
chiV = np.array([2 * G[classes[c][0]][0] for c in range(nc)])
vi = [i for i in range(nc) if degs[i] == 2 and np.allclose(chars[i], chiV, atol=1e-6)][0]
deg_vec = np.array(degs, float)
slice_ok = np.allclose(Nt[vi].dot(deg_vec), 2 * deg_vec)
check("HYP.FUSION.01 [E]: the 2I fusion tensor N^k_ij is a nonneg-integer weighted "
"3-uniform hypergraph on 9 nodes (entries %s); its 2-dim-rep slice = the "
"affine E8 McKay graph (A.deg=2deg, top eig 2)"
% sorted(set(Nt.flatten().tolist())),
(Nt >= 0).all() and slice_ok)
# HYP.SMITH.01
rs = {nm: _rho(_star(lg)) for nm, lg in
[("A8", [7, 0, 0]), ("E6", [1, 2, 2]), ("E7", [1, 2, 3]), ("E8", [1, 2, 4]),
("Ê8", [1, 2, 5]), ("E10", [1, 2, 6])]}
check("HYP.SMITH.01 [E]: Smith -- rho<=2 selects (affine) ADE; rho(A8,E6,E7,E8)<2 "
"(%.3f,%.3f,%.3f,%.3f), rho(Ê8)=%.3f, rho(E10)=%.3f>2; E8 = maximal finite "
"exceptional in T(2,3,r)"
% (rs["A8"], rs["E6"], rs["E7"], rs["E8"], rs["Ê8"], rs["E10"]),
rs["E8"] < 2 < rs["E10"] and abs(rs["Ê8"] - 2) < 1e-9)
# HYP.CW.01: the LOCAL gate is decided by the DIFFUSION-generated witness (no global
# rho); the Perron eigenvector is compared only as an independent audit.
cw = {nm: _witness_localgate(_star(lg))[1] for nm, lg in
[("E8", [1, 2, 4]), ("Ê8", [1, 2, 5]), ("E10", [1, 2, 6])]}
def _perron_gate(lg):
Ad = _star(lg).astype(float)
m = np.abs(np.linalg.eigh(Ad)[1][:, -1]); m = m / m.min()
return bool(np.all(Ad @ m <= 2 * m + 1e-6))
audit_ok = all(_perron_gate(lg) == cw[nm] for nm, lg in
[("E8", [1, 2, 4]), ("Ê8", [1, 2, 5]), ("E10", [1, 2, 6])])
check("HYP.CW.01 [E]: Collatz-Wielandt -- the LOCAL gate (diffusion-generated "
"witness, A m <= 2m, no global rho in the decision): E8=%s, Ê8=%s, E10=%s; "
"the independent Perron-eigenvector audit agrees (%s)"
% (cw["E8"], cw["Ê8"], cw["E10"], audit_ok),
cw["E8"] and cw["Ê8"] and not cw["E10"] and audit_ok)
# HYP.AUTO.01: autonomous rule from the (2,3,*) seed -> E6,E7,E8,Ê8
seq = _grow([1, 2, 2])
expected = [(1, 2, 2), (1, 2, 3), (1, 2, 4), (1, 2, 5)]
m_aff, _ = _witness_localgate(_star([1, 2, 5]))
marks = sorted(np.round(m_aff).astype(int).tolist())
check("HYP.AUTO.01 [E*]: the autonomous rule (self-generated witness via "
"projectively-normalised diffusion + LOCAL gate A m<=2m, no global rho) runs "
"E6->E7->E8->Ê8 then stops (seq %s); critical witness = Kac marks %s -- exact "
"UNDER the (2,3,*) seed + grow-longest-leg policy"
% (seq, marks),
seq == expected and marks == [1, 2, 2, 3, 3, 4, 4, 5, 6])
# HYP.NEG.AN.01: path control, EXACT for all n (closed form rho(A_m)=2cos(pi/(m+1)))
an_ok = all(abs(_rho(_star([m - 1, 0, 0])) - 2 * np.cos(np.pi / (m + 1))) < 1e-6
for m in range(4, 12))
seq_a = _grow([1, 0, 0], cap=8)
check("HYP.NEG.AN.01 [E]: control -- A_n is a path with rho(A_n)=2cos(pi/(n+1))<2 "
"for ALL n (closed form verified, n=4..11; finite growth %d steps never "
"reaches E8). So rho<=2 selects the ADE world, NOT E8"
% len(seq_a), an_ok and (1, 2, 4) not in seq_a)
# HYP.NEG.DN.01: D_n control, EXACT for all n (rho(D_m)=2cos(pi/(2m-2)))
dn_ok = all(abs(_rho(_star([1, 1, m - 3])) - 2 * np.cos(np.pi / (2 * m - 2))) < 1e-6
for m in range(4, 12))
seq_d = _grow([1, 1, 2], cap=8)
check("HYP.NEG.DN.01 [E]: control -- D_n has rho(D_n)=2cos(pi/(2n-2))<2 for ALL n "
"(closed form verified, n=4..11; finite growth %d steps never reaches E8). "
"Only the exceptional (1,2,r) arm reaches E8"
% len(seq_d), dn_ok and (1, 2, 4) not in seq_d)
# HYP.SEED.01: the seed is the irreducible input = P2 (NOT derived)
check("HYP.SEED.01 [O]: the 3-arm (2,3,r) exceptional branch type is the SINGLE "
"irreducible input = TFPT's P2 / the icosahedral choice (the A_n and D_n "
"controls confirm rho<=2 alone is not enough) -- this is NOT 'P2 derived', it "
"is 'P2 = the minimal seed on which the local rule REACHES E8 as the last "
"finite point'", True)
return summary("v299 hypergraph bridge: the 3-arm (2,3,r) seed is the irreducible input on which a local witness-based rule REACHES E8 as the last finite point (E6->E7->E8->Ê8; A_n/D_n controls with exact rho=2cos formulas: rho<=2 selects ADE, not E8); a second network reading of the (2,3,5)/E8 core, NOT a derivation of P2")
if __name__ == "__main__":
raise SystemExit(1 if run() else 0)
Recorded source fingerprint
0c4ecc2112669c92c0c07652ac1cf3a0df4c034dab574add63539ed82ba64b24
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.