================================================================================
UNIVERSAL SVAE DIAGNOSTIC BATTERY
================================================================================
  Loading: /root/.cache/huggingface/hub/models--AbstractPhil--geolip-SVAE/snapshots/6288b7c09958f9bf9ad193c65301e55ecd6c8836/v13_imagenet256/checkpoints/best.pt
  Epoch: 18, MSE: 0.000063
  Config: {'V': 256, 'D': 16, 'patch_size': 16, 'hidden': 768, 'depth': 4, 'n_cross_layers': 2, 'target_cv': 0.2915, 'cv_weight': 0.3, 'boost': 0.5, 'sigma': 0.15, 'dataset': 'imagenet_256', 'lr': 0.0001}
  Params: 16,942,419
  Inferred img_size=256 from dataset='imagenet_256'
  Resolution: 256×256, samples_per_test: 16

================================================================================
IMAGE DATASET BATTERY (256×256, n=16)
================================================================================
  dataset                       MSE        std        min        max |     S0     SD  ratio  erank
----------------------------------------------------------------------------------------------------
  CIFAR-10→256             0.000006   0.000001   0.000004   0.000008 |  5.044  3.212   1.57  15.85
  MNIST→256                0.000034   0.000055   0.000001   0.000207 |  5.040  3.214   1.57  15.85
  TinyImageNet→256         0.000007   0.000002   0.000003   0.000013 |  5.050  3.206   1.58  15.85
  ImageNet-128→256         0.000007   0.000002   0.000003   0.000010 |  5.044  3.225   1.56  15.85
Resolving data files: 100%
 40/40 [00:00<00:00, 13618.98it/s]
Resolving data files: 100%
 40/40 [00:00<00:00, 15192.63it/s]
  ImageNet-256→256         0.000038   0.000039   0.000007   0.000167 |  5.047  3.226   1.56  15.85

================================================================================
NOISE TYPE BATTERY (256×256, n=16)
================================================================================
  type                      MSE        std |     S0     SD  ratio  erank | byte_acc   ±1_acc
----------------------------------------------------------------------------------------------------
  gaussian             0.576848   0.017680 |  5.109  3.314   1.54  15.88 |    1.74%    5.21%
  uniform              0.137512   0.004244 |  5.084  3.259   1.56  15.87 |    3.37%   10.06%
  uniform_scaled       0.941287   0.031382 |  5.118  3.322   1.54  15.88 |    1.38%    4.13%
  poisson              0.108758   0.170382 |  5.067  3.254   1.56  15.86 |    5.91%   17.51%
  pink                 0.000003   0.000002 |  5.052  3.219   1.57  15.85 |   96.09%  100.00%
  brown                0.075172   0.300667 |  5.060  3.223   1.57  15.85 |   89.53%   93.78%
  salt_pepper          4.969349   0.094197 |  5.115  3.238   1.58  15.86 |    0.51%    1.52%
  sparse               0.300673   0.011496 |  5.103  3.290   1.55  15.87 |    2.84%    8.46%
  block                0.030697   0.017401 |  5.082  3.267   1.56  15.86 |   12.18%   33.63%
  gradient             0.097683   0.000679 |  5.075  3.213   1.58  15.86 |    4.04%   12.11%
  checkerboard         0.034273   0.000904 |  5.065  3.260   1.55  15.86 |    6.84%   20.32%
  mixed                0.160843   0.072098 |  5.088  3.264   1.56  15.87 |    3.35%   10.03%
  structural           0.848700   1.072937 |  5.104  3.290   1.55  15.87 |    5.16%    9.11%
  cauchy               3.540637   0.068339 |  5.116  3.262   1.57  15.87 |    0.70%    2.10%
  exponential          0.526940   0.027094 |  5.113  3.313   1.54  15.88 |    1.91%    5.69%
  laplace              1.627721   0.049655 |  5.120  3.311   1.55  15.88 |    1.17%    3.40%

================================================================================
TEXT BYTE RECONSTRUCTION (256×256)
================================================================================

  In:  'Hello, world! This is a test of the geometric encoder.'
  Out: 'K`hi_41fkmiY)(N`gg4^`4P.gdnh3^\(nf[/^bkkemqdZ1]fbjdcb1'
  MSE: 0.000006  Byte: 3.7%

  In:  'The quick brown fox jumps over the lazy dog. 0123456789'
  Out: 'TcY/ing_]3\inof)ajg3dnmic4eoge1gb]2`cuh4]g[4"*-/134465.'
  MSE: 0.000006  Byte: 3.6%

  In:  'import torch; model = PatchSVAE(); output = model(x)'
  Out: 'gjlmlh7fjmc]<.ahbc_+6'M]lbcSRAA((4.crsnnd/3*aicab6b,'
  MSE: 0.000004  Byte: 7.7%

  In:  'E = mc² — Albert Einstein, theoretical physicist, 1905'
  Out: '>(5*]k��P���5@b_cod-F`komeh`3+gdbkmdpfa\\3dfsoh`fof3&,5.*'
  MSE: 0.000008  Byte: 1.8%

  In:  'To be, or not to be, that is the question. — Shakespeare'
  Out: 'U`3VX4.^_4egc6hf&[V1/cd]`6^d5dec'ejhkpii_8?���-Oc`ffnldakW'
  MSE: 0.000009  Byte: 3.4%

  In:  '∫₀^∞ e^(-x²) dx = √π/2 — Gaussian integral'
  Out: 'Ѝ�̃{r���7XT41j��7,Zh,=B����}7-6���2D]ompe`Z4`hneela['
  MSE: 0.000012  Byte: 5.6%

  In:  '01101000 01100101 01101100 01101100 01101111 — binary hello'
  Out: '/00//..-#,/////.-"-0///0/-#,00./-/+&,00,00,2@���([ehdlh6^aih'
  MSE: 0.000005  Byte: 0.0%

  In:  'SELECT * FROM models WHERE cv BETWEEN 0.20 AND 0.23;'
  Out: 'OFICDI%)%?OLF/biachf0KEFJB.Wb2;AQRFDC'--0-'<H@%,-122'
  MSE: 0.000003  Byte: 0.0%

================================================================================
PIECEMEAL 1024→256 TILED RECONSTRUCTION
================================================================================
  gaussian          : 16 tiles, MSE=0.569875
  uniform           : 16 tiles, MSE=0.136389
  pink              : 16 tiles, MSE=0.000006
  salt_pepper       : 16 tiles, MSE=4.967982
  cauchy            : 16 tiles, MSE=3.490826

================================================================================
SIGNAL ENERGY SURVIVAL
================================================================================
  source                   survival     SNR_dB     orig_E    recon_E
----------------------------------------------------------------------
  CIFAR-10→256               100.0%      51.6dB     0.9119     0.9118
  MNIST→256                   99.9%      43.7dB     0.7910     0.7905
  TinyImageNet→256           100.0%      51.1dB     0.8605     0.8606
  ImageNet-128→256           100.0%      51.7dB     1.0926     1.0924
Resolving data files: 100%
 40/40 [00:00<00:00, 12296.41it/s]
Resolving data files: 100%
 40/40 [00:00<00:00, 13358.72it/s]
  ImageNet-256→256           100.0%      44.9dB     1.1830     1.1828
  noise/gaussian               83.7%       2.4dB     0.9989     0.8358
  noise/pink                   95.3%      25.7dB     2.0899     1.9916
  noise/salt_pepper            71.9%      -0.9dB     4.0104     2.8823
  noise/cauchy                 86.1%      -0.8dB     2.9567     2.5454

================================================================================
ALPHA PROFILE
================================================================================

  Layer 0: mean=0.02945  max=0.03004  min=0.02855  std=0.000477
    α[ 0]: 0.02855  ███████████████████████████████████████████████
    α[ 1]: 0.02976  █████████████████████████████████████████████████
    α[ 2]: 0.02955  █████████████████████████████████████████████████
    α[ 3]: 0.02956  █████████████████████████████████████████████████
    α[ 4]: 0.02927  ████████████████████████████████████████████████
    α[ 5]: 0.03004  █████████████████████████████████████████████████
    α[ 6]: 0.02859  ███████████████████████████████████████████████
    α[ 7]: 0.02868  ███████████████████████████████████████████████
    α[ 8]: 0.02932  ████████████████████████████████████████████████
    α[ 9]: 0.03001  █████████████████████████████████████████████████
    α[10]: 0.03003  █████████████████████████████████████████████████
    α[11]: 0.02966  █████████████████████████████████████████████████
    α[12]: 0.02968  █████████████████████████████████████████████████
    α[13]: 0.02954  █████████████████████████████████████████████████
    α[14]: 0.02940  ████████████████████████████████████████████████
    α[15]: 0.02950  █████████████████████████████████████████████████

  Layer 1: mean=0.02962  max=0.03055  min=0.02902  std=0.000476
    α[ 0]: 0.03055  █████████████████████████████████████████████████
    α[ 1]: 0.02933  ███████████████████████████████████████████████
    α[ 2]: 0.02952  ████████████████████████████████████████████████
    α[ 3]: 0.02950  ████████████████████████████████████████████████
    α[ 4]: 0.02980  ████████████████████████████████████████████████
    α[ 5]: 0.02902  ███████████████████████████████████████████████
    α[ 6]: 0.03044  █████████████████████████████████████████████████
    α[ 7]: 0.03034  █████████████████████████████████████████████████
    α[ 8]: 0.02972  ████████████████████████████████████████████████
    α[ 9]: 0.02903  ███████████████████████████████████████████████
    α[10]: 0.02902  ███████████████████████████████████████████████
    α[11]: 0.02939  ████████████████████████████████████████████████
    α[12]: 0.02939  ████████████████████████████████████████████████
    α[13]: 0.02952  ████████████████████████████████████████████████
    α[14]: 0.02967  ████████████████████████████████████████████████
    α[15]: 0.02961  ████████████████████████████████████████████████

================================================================================
COMPRESSION METRICS
================================================================================
  Input:   256×256×3 = 196,608 values
  Latent:  16×256 = 4,096 omega tokens
  Ratio:   48.0:1
  8-bit: input=192.0KB  latent=4.0KB  ratio=48.0:1
  16-bit: input=384.0KB  latent=8.0KB  ratio=48.0:1
  32-bit: input=768.0KB  latent=16.0KB  ratio=48.0:1

================================================================================
RECONSTRUCTION GRID
================================================================================
Resolving data files: 100%
 40/40 [00:00<00:00, 13764.23it/s]
Resolving data files: 100%
 40/40 [00:00<00:00, 15695.78it/s]
  Saved: universal_diagnostic_grid.png

  Results: diagnostic_v13_imagenet256.json

================================================================================
DIAGNOSTIC COMPLETE
================================================================================