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_0.0005.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: 0.09657791256904602
Inter Cos: 0.11771981418132782
Norm Quadratic Average: 64.1571044921875
Nearest Class Center Accuracy: 0.8092

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.12315083295106888
Inter Cos: 0.15018528699874878
Norm Quadratic Average: 84.9493408203125
Nearest Class Center Accuracy: 0.8199

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.12539318203926086
Inter Cos: 0.1589694619178772
Norm Quadratic Average: 138.1958770751953
Nearest Class Center Accuracy: 0.8262833333333334

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1945667266845703
Inter Cos: 0.1807381808757782
Norm Quadratic Average: 110.09539794921875
Nearest Class Center Accuracy: 0.8805166666666666

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.22094067931175232
Inter Cos: 0.1897362470626831
Norm Quadratic Average: 102.70836639404297
Nearest Class Center Accuracy: 0.9084333333333333

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.24233244359493256
Inter Cos: 0.1902221143245697
Norm Quadratic Average: 82.80177307128906
Nearest Class Center Accuracy: 0.9302166666666667

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2679072618484497
Inter Cos: 0.18592076003551483
Norm Quadratic Average: 56.38371276855469
Nearest Class Center Accuracy: 0.9476166666666667

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2938966751098633
Inter Cos: 0.15308965742588043
Norm Quadratic Average: 22.073192596435547
Nearest Class Center Accuracy: 0.9688

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4714055061340332
Inter Cos: 0.24629364907741547
Norm Quadratic Average: 13.360840797424316
Nearest Class Center Accuracy: 0.9835666666666667

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5813174247741699
Inter Cos: 0.28692519664764404
Norm Quadratic Average: 12.262011528015137
Nearest Class Center Accuracy: 0.9882333333333333

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6481308937072754
Inter Cos: 0.2871759533882141
Norm Quadratic Average: 12.840946197509766
Nearest Class Center Accuracy: 0.9930833333333333

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6963443756103516
Inter Cos: 0.3485286235809326
Norm Quadratic Average: 9.067815780639648
Nearest Class Center Accuracy: 0.99285

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8563219904899597
Inter Cos: 0.3945372700691223
Norm Quadratic Average: 8.151540756225586
Nearest Class Center Accuracy: 0.9962166666666666

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8822996616363525
Inter Cos: 0.3444966971874237
Norm Quadratic Average: 8.115747451782227
Nearest Class Center Accuracy: 0.9972166666666666

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8966866135597229
Inter Cos: 0.3209066689014435
Norm Quadratic Average: 7.634598731994629
Nearest Class Center Accuracy: 0.99755

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 14.706413269042969
Linear Weight Rank: 4031
Intra Cos: 0.9126293659210205
Inter Cos: 0.2641840875148773
Norm Quadratic Average: 40.27093505859375
Nearest Class Center Accuracy: 0.99835

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 6.338138103485107
Linear Weight Rank: 3667
Intra Cos: 0.9252328872680664
Inter Cos: 0.2724952697753906
Norm Quadratic Average: 33.00774002075195
Nearest Class Center Accuracy: 0.9989

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.312424421310425
Linear Weight Rank: 10
Intra Cos: 0.9282694458961487
Inter Cos: 0.2381402850151062
Norm Quadratic Average: 26.652591705322266
Nearest Class Center Accuracy: 0.9993

Output Layer:
Intra Cos: 0.9557669758796692
Inter Cos: 0.29512086510658264
Norm Quadratic Average: 23.907028198242188
Nearest Class Center Accuracy: 0.9999166666666667

Test Set:
Average Loss: 0.025226062105850725
Accuracy: 0.9937
NC1 Within Class Collapse: 0.7918334603309631
NC2 Equinorm: Features: 0.10243681818246841, Weights: 0.027617601677775383
NC2 Equiangle: Features: 0.2450030008951823, Weights: 0.15617941750420464
NC3 Self-Duality: 0.13270153105258942
NC4 NCC Mismatch: 0.0031999999999999806

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: 0.10789725184440613
Inter Cos: 0.12962490320205688
Norm Quadratic Average: 64.3337631225586
Nearest Class Center Accuracy: 0.8221

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1376909613609314
Inter Cos: 0.16454188525676727
Norm Quadratic Average: 84.94881439208984
Nearest Class Center Accuracy: 0.8329

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.13990649580955505
Inter Cos: 0.17407165467739105
Norm Quadratic Average: 138.25115966796875
Nearest Class Center Accuracy: 0.8396

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2115194946527481
Inter Cos: 0.1964542716741562
Norm Quadratic Average: 110.01067352294922
Nearest Class Center Accuracy: 0.8911

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.23867632448673248
Inter Cos: 0.2055826485157013
Norm Quadratic Average: 102.68479919433594
Nearest Class Center Accuracy: 0.9186

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.25891947746276855
Inter Cos: 0.20640282332897186
Norm Quadratic Average: 82.87739562988281
Nearest Class Center Accuracy: 0.9375

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.28333592414855957
Inter Cos: 0.18802708387374878
Norm Quadratic Average: 56.526817321777344
Nearest Class Center Accuracy: 0.9513

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3058183491230011
Inter Cos: 0.16710977256298065
Norm Quadratic Average: 22.15264320373535
Nearest Class Center Accuracy: 0.9713

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.489967942237854
Inter Cos: 0.261637419462204
Norm Quadratic Average: 13.428018569946289
Nearest Class Center Accuracy: 0.9801

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5999056100845337
Inter Cos: 0.3067859411239624
Norm Quadratic Average: 12.338582038879395
Nearest Class Center Accuracy: 0.9833

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6563522219657898
Inter Cos: 0.3028266131877899
Norm Quadratic Average: 12.931870460510254
Nearest Class Center Accuracy: 0.9871

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6956404447555542
Inter Cos: 0.3473019003868103
Norm Quadratic Average: 9.130354881286621
Nearest Class Center Accuracy: 0.985

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8560740947723389
Inter Cos: 0.38917139172554016
Norm Quadratic Average: 8.222021102905273
Nearest Class Center Accuracy: 0.9888

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.887282133102417
Inter Cos: 0.3465607762336731
Norm Quadratic Average: 8.188837051391602
Nearest Class Center Accuracy: 0.9893

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8996097445487976
Inter Cos: 0.3224332928657532
Norm Quadratic Average: 7.7016167640686035
Nearest Class Center Accuracy: 0.9902

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 14.706413269042969
Linear Weight Rank: 4031
Intra Cos: 0.9141802787780762
Inter Cos: 0.25981494784355164
Norm Quadratic Average: 40.595191955566406
Nearest Class Center Accuracy: 0.9911

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 6.338138103485107
Linear Weight Rank: 3667
Intra Cos: 0.9250203967094421
Inter Cos: 0.26736441254615784
Norm Quadratic Average: 33.27219009399414
Nearest Class Center Accuracy: 0.9918

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.312424421310425
Linear Weight Rank: 10
Intra Cos: 0.926390528678894
Inter Cos: 0.2446591556072235
Norm Quadratic Average: 26.868505477905273
Nearest Class Center Accuracy: 0.9922

Output Layer:
Intra Cos: 0.946631133556366
Inter Cos: 0.2964126467704773
Norm Quadratic Average: 24.09356689453125
Nearest Class Center Accuracy: 0.9926

