File size: 13,231 Bytes
396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 526518a 396b9e6 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 | """
SVAE β Structural Binding Constant
=====================================
Matrix (V, 24): V rows in D=24 space.
At D=24, CV β 0.29154 BY CONSTRUCTION β no loss needed.
The sweep proved it:
V=200, D=24 β CV=0.2914
V=1024, D=24 β CV=0.2916
V=1992, D=24 β CV=0.2911
V is irrelevant. D determines CV.
The encoder produces a (V, 24) matrix.
The rows ARE an embedding: V tokens in D=24 space.
Their CV is ~0.29 by the dimensional law.
The SVD decomposes this embedding into its spectral structure.
The decoder reconstructs from the decomposition.
No CV loss. Monitor only. The geometry is inherent.
pip install "git+https://github.com/AbstractEyes/geolip-core.git"
"""
import torch
import torch.nn as nn
import torch.nn.functional as F
import torchvision
import torchvision.transforms as T
import math
try:
from geolip_core.linalg import svd as geolip_svd
HAS_GEOLIP = True
print("Using geolip-core SVD (Gram + eigh)")
except ImportError:
HAS_GEOLIP = False
print("geolip-core not found, fallback to torch.svd_lowrank")
# ββ CM for monitoring (not loss) ββ
def cayley_menger_vol2(points):
B, N, D = points.shape
gram = torch.bmm(points, points.transpose(1, 2))
norms = torch.diagonal(gram, dim1=1, dim2=2)
d2 = F.relu(norms.unsqueeze(2) + norms.unsqueeze(1) - 2 * gram)
cm = torch.zeros(B, N + 1, N + 1, device=points.device, dtype=points.dtype)
cm[:, 0, 1:] = 1.0
cm[:, 1:, 0] = 1.0
cm[:, 1:, 1:] = d2
k = N - 1
sign = (-1.0) ** (k + 1)
fact = math.factorial(k)
return sign * torch.linalg.det(cm.float()).to(points.dtype) / ((2 ** k) * (fact ** 2))
def cv_of(emb, n_samples=200):
if emb.dim() != 2 or emb.shape[0] < 5:
return 0.0
N, D = emb.shape
pool = min(N, 512)
indices = torch.stack([torch.randperm(pool, device=emb.device)[:5] for _ in range(n_samples)])
vol2 = cayley_menger_vol2(emb[:pool][indices])
valid = vol2 > 1e-20
if valid.sum() < 10:
return 0.0
vols = vol2[valid].sqrt()
return (vols.std() / (vols.mean() + 1e-8)).item()
BINDING_CONSTANT = 0.29154
# ββ Data ββ
def get_cifar10(batch_size=256):
transform = T.Compose([
T.ToTensor(),
T.Normalize((0.4914, 0.4822, 0.4465), (0.2470, 0.2435, 0.2616)),
])
train_ds = torchvision.datasets.CIFAR10(root='./data', train=True, download=True, transform=transform)
test_ds = torchvision.datasets.CIFAR10(root='./data', train=False, download=True, transform=transform)
train_loader = torch.utils.data.DataLoader(train_ds, batch_size=batch_size, shuffle=True, num_workers=2)
test_loader = torch.utils.data.DataLoader(test_ds, batch_size=batch_size, shuffle=False, num_workers=2)
return train_loader, test_loader
# ββ SVAE ββ
class SVAE(nn.Module):
def __init__(self, matrix_v=48, D=24):
"""
matrix_v: number of rows (vocabulary size of the implicit embedding)
D: embedding dimension = number of singular values = 24 for binding constant
"""
super().__init__()
self.matrix_v = matrix_v # V β number of embedding rows
self.D = D # D β embedding dimension
self.img_dim = 3 * 32 * 32
self.mat_dim = matrix_v * D
self.encoder = nn.Sequential(
nn.Linear(self.img_dim, 512),
nn.GELU(),
nn.Linear(512, 512),
nn.GELU(),
nn.Linear(512, self.mat_dim),
)
self.decoder = nn.Sequential(
nn.Linear(self.mat_dim, 512),
nn.GELU(),
nn.Linear(512, 512),
nn.GELU(),
nn.Linear(512, self.img_dim),
)
def encode(self, images):
B = images.shape[0]
M = self.encoder(images.reshape(B, -1)).reshape(B, self.matrix_v, self.D)
if HAS_GEOLIP:
U, S, Vh = geolip_svd(M)
else:
U, S, V = torch.svd_lowrank(M, q=self.D)
Vh = V.transpose(1, 2)
return {
'U': U, 'S': S, 'Vt': Vh,
'M': M, # the embedding matrix β rows are V points in D=24
}
def decode_from_svd(self, U, S, Vt):
B = U.shape[0]
M_hat = torch.bmm(U * S.unsqueeze(1), Vt)
return self.decoder(M_hat.reshape(B, -1)).reshape(B, 3, 32, 32)
def forward(self, images):
svd = self.encode(images)
recon = self.decode_from_svd(svd['U'], svd['S'], svd['Vt'])
return {'recon': recon, 'svd': svd}
@staticmethod
def effective_rank(S):
p = S / (S.sum(-1, keepdim=True) + 1e-8)
p = p.clamp(min=1e-8)
return (-(p * p.log()).sum(-1)).exp()
# ββ Training ββ
def train(epochs=50, lr=1e-3, device='cuda'):
device = torch.device(device if torch.cuda.is_available() else 'cpu')
train_loader, test_loader = get_cifar10(batch_size=256)
D = 24
V = 48
model = SVAE(matrix_v=V, D=D).to(device)
opt = torch.optim.Adam(model.parameters(), lr=lr)
sched = torch.optim.lr_scheduler.CosineAnnealingLR(opt, T_max=epochs)
total_params = sum(p.numel() for p in model.parameters())
print(f"SVAE β Structural Binding Constant")
print(f" Matrix: ({V}, {D}) β {V} rows in D={D} space")
print(f" Expected row CV β {BINDING_CONSTANT} (no loss, by construction)")
print(f" SVD: {'geolip-core' if HAS_GEOLIP else 'torch.svd_lowrank'}")
print(f" Compression: {model.img_dim} β {D} ({model.img_dim // D}:1)")
print(f" Params: {total_params:,}")
print("=" * 85)
print(f"{'ep':>3} | {'loss':>7} {'recon':>7} | "
f"{'t_recon':>7} | "
f"{'S0':>6} {'SD':>6} {'ratio':>6} {'erank':>6} | "
f"{'row_cv':>7} {'Ξbc':>7}")
print("-" * 85)
for epoch in range(1, epochs + 1):
model.train()
total_loss, n = 0, 0
for images, labels in train_loader:
images = images.to(device)
opt.zero_grad()
out = model(images)
loss = F.mse_loss(out['recon'], images)
loss.backward()
opt.step()
total_loss += loss.item() * len(images)
n += len(images)
sched.step()
if epoch % 2 == 0 or epoch <= 3:
model.eval()
test_recon, test_n = 0, 0
test_S = None
test_erank = 0
row_cvs = []
nb = 0
with torch.no_grad():
for images, labels in test_loader:
images = images.to(device)
out = model(images)
test_recon += F.mse_loss(out['recon'], images).item() * len(images)
test_n += len(images)
test_erank += model.effective_rank(out['svd']['S']).mean().item()
# CV of matrix rows: each M[i] is (V, D) β V points in D=24
# Sample a few to keep it fast
if nb < 5:
for b in range(min(4, len(images))):
row_cvs.append(cv_of(out['svd']['M'][b]))
if test_S is None:
test_S = out['svd']['S'].mean(0).cpu()
else:
test_S += out['svd']['S'].mean(0).cpu()
nb += 1
test_erank /= nb
test_S /= nb
ratio = (test_S[0] / (test_S[-1] + 1e-8)).item()
mean_row_cv = sum(row_cvs) / len(row_cvs) if row_cvs else 0
delta_bc = abs(mean_row_cv - BINDING_CONSTANT)
print(f"{epoch:3d} | {total_loss/n:7.4f} {total_loss/n:7.4f} | "
f"{test_recon/test_n:7.4f} | "
f"{test_S[0]:6.3f} {test_S[-1]:6.3f} {ratio:6.2f} "
f"{test_erank:6.2f} | "
f"{mean_row_cv:7.4f} {delta_bc:7.4f}")
# ββ Final Analysis ββ
print()
print("=" * 85)
print("FINAL ANALYSIS")
print("=" * 85)
model.eval()
all_S, all_recon_err, all_labels = [], [], []
all_row_cvs = []
with torch.no_grad():
for images, labels in test_loader:
images = images.to(device)
out = model(images)
all_S.append(out['svd']['S'].cpu())
all_recon_err.append(
F.mse_loss(out['recon'], images, reduction='none')
.mean(dim=(1, 2, 3)).cpu())
all_labels.append(labels.cpu())
# Row CV for a sample of images
for b in range(min(8, len(images))):
all_row_cvs.append(cv_of(out['svd']['M'][b]))
all_S = torch.cat(all_S)
all_recon_err = torch.cat(all_recon_err)
all_labels = torch.cat(all_labels)
erank = model.effective_rank(all_S)
mean_row_cv = sum(all_row_cvs) / len(all_row_cvs)
print(f"\n Architecture: ({V}, {D}) β {V} rows Γ D={D}")
print(f" Recon MSE: {all_recon_err.mean():.6f} Β± {all_recon_err.std():.6f}")
print(f" Effective rank: {erank.mean():.2f} Β± {erank.std():.2f}")
print(f"\n Row CV (matrix rows as D={D} embedding):")
print(f" Measured: {mean_row_cv:.4f}")
print(f" Target: {BINDING_CONSTANT}")
print(f" Delta: {abs(mean_row_cv - BINDING_CONSTANT):.4f}")
print(f" {'β AT BINDING CONSTANT' if abs(mean_row_cv - BINDING_CONSTANT) < 0.01 else 'β Not at binding constant'}")
# Spectrum profile
S_mean = all_S.mean(0)
total_energy = (S_mean ** 2).sum()
print(f"\n Singular value profile:")
cumulative = 0
for i in range(len(S_mean)):
e = (S_mean[i] ** 2).item()
cumulative += e
pct = cumulative / total_energy * 100
bar = "β" * int(S_mean[i].item() * 30 / (S_mean[0].item() + 1e-8))
print(f" S[{i:2d}]: {S_mean[i]:8.3f} cum={pct:5.1f}% {bar}")
# Per-class
cifar_names = ['plane', 'car', 'bird', 'cat', 'deer',
'dog', 'frog', 'horse', 'ship', 'truck']
print(f"\n Per-class:")
print(f" {'class':>6} {'recon':>8} {'erank':>6} {'S0':>7} {'SD':>7} {'ratio':>6}")
for c in range(10):
mask = all_labels == c
rc = all_recon_err[mask].mean().item()
er = erank[mask].mean().item()
s0 = all_S[mask, 0].mean().item()
sd = all_S[mask, -1].mean().item()
ratio = s0 / (sd + 1e-8)
print(f" {cifar_names[c]:>6} {rc:8.6f} {er:6.2f} {s0:7.3f} {sd:7.3f} {ratio:6.2f}")
# Cross-class spectral variance
class_S_means = torch.stack([all_S[all_labels == c].mean(0) for c in range(10)])
s_var = class_S_means.std(0)
print(f"\n Cross-class S variance (top 5 most discriminative):")
_, top_idx = s_var.topk(5)
for idx in top_idx:
i = idx.item()
print(f" S[{i:2d}]: var={s_var[i]:.4f}")
# ββ Reconstruction grid ββ
print(f"\n Saving reconstruction grid...")
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
mean_t = torch.tensor([0.4914, 0.4822, 0.4465]).reshape(1, 3, 1, 1).to(device)
std_t = torch.tensor([0.2470, 0.2435, 0.2616]).reshape(1, 3, 1, 1).to(device)
model.eval()
with torch.no_grad():
images, labels = next(iter(test_loader))
images = images.to(device)
out = model(images)
selected_idx = []
for c in range(10):
class_idx = (labels == c).nonzero(as_tuple=True)[0]
selected_idx.extend(class_idx[:2].tolist())
orig = images[selected_idx]
U = out['svd']['U'][selected_idx]
S = out['svd']['S'][selected_idx]
Vt = out['svd']['Vt'][selected_idx]
mode_counts = [1, 4, 8, 16, D]
mode_counts = list(dict.fromkeys([m for m in mode_counts if m <= D]))
prog_recons = []
for n_modes in mode_counts:
r = model.decode_from_svd(U[:, :, :n_modes], S[:, :n_modes], Vt[:, :n_modes, :])
prog_recons.append(r)
def denorm(t):
return (t * std_t + mean_t).clamp(0, 1).cpu()
n_samples = len(selected_idx)
n_cols = 2 + len(mode_counts)
fig, axes = plt.subplots(n_samples, n_cols, figsize=(n_cols * 1.5, n_samples * 1.5))
col_titles = ['Original'] + [f'{m} mode{"s" if m > 1 else ""}' for m in mode_counts] + ['|Error|Γ5']
for i in range(n_samples):
axes[i, 0].imshow(denorm(orig[i:i+1])[0].permute(1, 2, 0).numpy())
for j, r in enumerate(prog_recons):
axes[i, j+1].imshow(denorm(r[i:i+1])[0].permute(1, 2, 0).numpy())
err_col = 1 + len(prog_recons)
diff = (denorm(orig[i:i+1]) - denorm(prog_recons[-1][i:i+1])).abs() * 5
axes[i, err_col].imshow(diff.clamp(0, 1)[0].permute(1, 2, 0).numpy())
c = labels[selected_idx[i]].item()
axes[i, 0].set_ylabel(cifar_names[c], fontsize=8, rotation=0, labelpad=35)
for j, title in enumerate(col_titles):
axes[0, j].set_title(title, fontsize=8)
for ax in axes.flat:
ax.axis('off')
plt.tight_layout()
plt.savefig('/content/svae_recon_grid.png', dpi=200, bbox_inches='tight')
print(f" Saved to /content/svae_recon_grid.png")
try:
plt.show()
except:
pass
plt.close()
if __name__ == "__main__":
train() |