Model save path: ./New_Models/bn_False_dataset_MNIST_epochs_200_lr_0.01_model_type_vgg19_rand_seed_314159_test_samples_None_train_samples_None_weight_decay_0.01.pth.tar
Training Set:
Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.09116753935813904
Inter Cos: 0.10967151075601578
Norm Quadratic Average: 23.567678451538086
Nearest Class Center Accuracy: 0.8079833333333334

Layer 1: Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.11942823231220245
Inter Cos: 0.1436183899641037
Norm Quadratic Average: 66.71934509277344
Nearest Class Center Accuracy: 0.7987333333333333

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.14187215268611908
Inter Cos: 0.18452458083629608
Norm Quadratic Average: 130.50759887695312
Nearest Class Center Accuracy: 0.7738833333333334

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1468943953514099
Inter Cos: 0.19922825694084167
Norm Quadratic Average: 273.39202880859375
Nearest Class Center Accuracy: 0.7625833333333333

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1612570583820343
Inter Cos: 0.2013588398694992
Norm Quadratic Average: 226.80056762695312
Nearest Class Center Accuracy: 0.7582666666666666

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.17034363746643066
Inter Cos: 0.20307935774326324
Norm Quadratic Average: 208.70068359375
Nearest Class Center Accuracy: 0.7711

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.18137933313846588
Inter Cos: 0.2097231149673462
Norm Quadratic Average: 204.3263702392578
Nearest Class Center Accuracy: 0.8196

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2146572321653366
Inter Cos: 0.22606626152992249
Norm Quadratic Average: 181.32188415527344
Nearest Class Center Accuracy: 0.8794666666666666

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.18295660614967346
Inter Cos: 0.2440025955438614
Norm Quadratic Average: 75.9161148071289
Nearest Class Center Accuracy: 0.8588666666666667

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.17488087713718414
Inter Cos: 0.30383655428886414
Norm Quadratic Average: 42.8765869140625
Nearest Class Center Accuracy: 0.7841833333333333

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2702217400074005
Inter Cos: 0.3626406788825989
Norm Quadratic Average: 36.97396469116211
Nearest Class Center Accuracy: 0.8364333333333334

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4099508821964264
Inter Cos: 0.406297504901886
Norm Quadratic Average: 36.39658737182617
Nearest Class Center Accuracy: 0.90975

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.44111254811286926
Inter Cos: 0.33709537982940674
Norm Quadratic Average: 20.933137893676758
Nearest Class Center Accuracy: 0.8923666666666666

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5936958193778992
Inter Cos: 0.4948633313179016
Norm Quadratic Average: 15.942063331604004
Nearest Class Center Accuracy: 0.90825

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.681879997253418
Inter Cos: 0.563567578792572
Norm Quadratic Average: 17.047815322875977
Nearest Class Center Accuracy: 0.9458

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7545254826545715
Inter Cos: 0.6214026808738708
Norm Quadratic Average: 18.82347869873047
Nearest Class Center Accuracy: 0.9661666666666666

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 1.498976707458496
Linear Weight Rank: 7
Intra Cos: 0.8139351010322571
Inter Cos: 0.5773128867149353
Norm Quadratic Average: 80.31842041015625
Nearest Class Center Accuracy: 0.9798333333333333

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 1.5061829090118408
Linear Weight Rank: 2723
Intra Cos: 0.8789684176445007
Inter Cos: 0.511726438999176
Norm Quadratic Average: 51.65919494628906
Nearest Class Center Accuracy: 0.99005

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 1.4967595338821411
Linear Weight Rank: 9
Intra Cos: 0.9021835923194885
Inter Cos: 0.42948463559150696
Norm Quadratic Average: 29.375707626342773
Nearest Class Center Accuracy: 0.9933666666666666

Output Layer:
Intra Cos: 0.9498628973960876
Inter Cos: 0.48786985874176025
Norm Quadratic Average: 18.060792922973633
Nearest Class Center Accuracy: 0.9951166666666666

Test Set:
Average Loss: 0.04837594205290079
Accuracy: 0.9857
NC1 Within Class Collapse: 1.3971939086914062
NC2 Equinorm: Features: 0.09948637336492538, Weights: 0.04442160949110985
NC2 Equiangle: Features: 0.30089717441134983, Weights: 0.19597839779324003
NC3 Self-Duality: 0.09736334532499313
NC4 NCC Mismatch: 0.00770000000000004

Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10219582915306091
Inter Cos: 0.12048851698637009
Norm Quadratic Average: 23.595197677612305
Nearest Class Center Accuracy: 0.8229

Layer 1: Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.13116680085659027
Inter Cos: 0.1572125256061554
Norm Quadratic Average: 67.1442642211914
Nearest Class Center Accuracy: 0.8148

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.15641117095947266
Inter Cos: 0.20206020772457123
Norm Quadratic Average: 131.20277404785156
Nearest Class Center Accuracy: 0.7946

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.161960631608963
Inter Cos: 0.2187100499868393
Norm Quadratic Average: 274.8509826660156
Nearest Class Center Accuracy: 0.7838

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1718568354845047
Inter Cos: 0.22334828972816467
Norm Quadratic Average: 227.4340362548828
Nearest Class Center Accuracy: 0.7815

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.18256600201129913
Inter Cos: 0.2255421280860901
Norm Quadratic Average: 209.2810821533203
Nearest Class Center Accuracy: 0.7933

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.19321416318416595
Inter Cos: 0.23281829059123993
Norm Quadratic Average: 205.10426330566406
Nearest Class Center Accuracy: 0.8407

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.22607843577861786
Inter Cos: 0.23746143281459808
Norm Quadratic Average: 182.47727966308594
Nearest Class Center Accuracy: 0.8921

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.19243033230304718
Inter Cos: 0.2435808926820755
Norm Quadratic Average: 76.35549926757812
Nearest Class Center Accuracy: 0.8719

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.17741592228412628
Inter Cos: 0.3253166079521179
Norm Quadratic Average: 43.09415817260742
Nearest Class Center Accuracy: 0.8045

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2705671787261963
Inter Cos: 0.38825297355651855
Norm Quadratic Average: 37.203956604003906
Nearest Class Center Accuracy: 0.8468

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.41126102209091187
Inter Cos: 0.4257326126098633
Norm Quadratic Average: 36.72523880004883
Nearest Class Center Accuracy: 0.911

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.44871288537979126
Inter Cos: 0.34349650144577026
Norm Quadratic Average: 21.140213012695312
Nearest Class Center Accuracy: 0.8967

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5988121032714844
Inter Cos: 0.49972662329673767
Norm Quadratic Average: 16.120563507080078
Nearest Class Center Accuracy: 0.9081

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6826975345611572
Inter Cos: 0.5647779107093811
Norm Quadratic Average: 17.248760223388672
Nearest Class Center Accuracy: 0.939

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7524840831756592
Inter Cos: 0.619895339012146
Norm Quadratic Average: 19.0518798828125
Nearest Class Center Accuracy: 0.9568

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 1.498976707458496
Linear Weight Rank: 7
Intra Cos: 0.8114517331123352
Inter Cos: 0.576454758644104
Norm Quadratic Average: 81.39582824707031
Nearest Class Center Accuracy: 0.9705

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 1.5061829090118408
Linear Weight Rank: 2723
Intra Cos: 0.8720349669456482
Inter Cos: 0.510543167591095
Norm Quadratic Average: 52.36581039428711
Nearest Class Center Accuracy: 0.9804

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 1.4967595338821411
Linear Weight Rank: 9
Intra Cos: 0.8888404369354248
Inter Cos: 0.42806780338287354
Norm Quadratic Average: 29.772470474243164
Nearest Class Center Accuracy: 0.9828

Output Layer:
Intra Cos: 0.9306880235671997
Inter Cos: 0.4868552088737488
Norm Quadratic Average: 18.31671905517578
Nearest Class Center Accuracy: 0.9842

