Model save path: ./New_Models/bn_True_dataset_MNIST_epochs_200_lr_0.01_model_type_vgg19_rand_seed_265358_test_samples_None_train_samples_None_weight_decay_0.005.pth.tar
Training Set:
Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.09116753190755844
Inter Cos: 0.10967151820659637
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: 0.05953482538461685
Inter Cos: 0.07724731415510178
Norm Quadratic Average: 2.5180561542510986
Nearest Class Center Accuracy: 0.8103166666666667

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10403826832771301
Inter Cos: 0.10411425679922104
Norm Quadratic Average: 1.550970435142517
Nearest Class Center Accuracy: 0.8736666666666667

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.09649129211902618
Inter Cos: 0.09875360131263733
Norm Quadratic Average: 1.2041758298873901
Nearest Class Center Accuracy: 0.87965

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.17555326223373413
Inter Cos: 0.12122868001461029
Norm Quadratic Average: 0.8248242735862732
Nearest Class Center Accuracy: 0.9384

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2285931408405304
Inter Cos: 0.13446082174777985
Norm Quadratic Average: 0.6099066138267517
Nearest Class Center Accuracy: 0.9634166666666667

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3028185963630676
Inter Cos: 0.1447225958108902
Norm Quadratic Average: 0.5118808150291443
Nearest Class Center Accuracy: 0.97395

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.34845244884490967
Inter Cos: 0.1334673911333084
Norm Quadratic Average: 0.4428272247314453
Nearest Class Center Accuracy: 0.9783833333333334

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4211931526660919
Inter Cos: 0.15202568471431732
Norm Quadratic Average: 0.28532886505126953
Nearest Class Center Accuracy: 0.9936833333333334

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6554413437843323
Inter Cos: 0.2584613561630249
Norm Quadratic Average: 0.1909622848033905
Nearest Class Center Accuracy: 0.9986833333333334

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8065022826194763
Inter Cos: 0.26368623971939087
Norm Quadratic Average: 0.19127078354358673
Nearest Class Center Accuracy: 0.9995166666666667

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8452542424201965
Inter Cos: 0.1625901758670807
Norm Quadratic Average: 0.1881645917892456
Nearest Class Center Accuracy: 0.99995

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9083694219589233
Inter Cos: 0.17060089111328125
Norm Quadratic Average: 0.22724959254264832
Nearest Class Center Accuracy: 0.99995

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9865573048591614
Inter Cos: 0.0352962464094162
Norm Quadratic Average: 0.273335337638855
Nearest Class Center Accuracy: 0.9999666666666667

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9969280958175659
Inter Cos: 0.003197215497493744
Norm Quadratic Average: 0.534269392490387
Nearest Class Center Accuracy: 0.9999666666666667

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9985800385475159
Inter Cos: 0.040714774280786514
Norm Quadratic Average: 1.1039425134658813
Nearest Class Center Accuracy: 0.9999666666666667

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.144974708557129
Linear Weight Rank: 10
Intra Cos: 0.9991263747215271
Inter Cos: 0.1259474903345108
Norm Quadratic Average: 25.064592361450195
Nearest Class Center Accuracy: 0.9999666666666667

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.1474530696868896
Linear Weight Rank: 1426
Intra Cos: 0.9992530941963196
Inter Cos: 0.1788238286972046
Norm Quadratic Average: 17.20442008972168
Nearest Class Center Accuracy: 0.9999666666666667

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.1475274562835693
Linear Weight Rank: 9
Intra Cos: 0.9993184804916382
Inter Cos: 0.1570516675710678
Norm Quadratic Average: 12.043542861938477
Nearest Class Center Accuracy: 0.9999666666666667

Output Layer:
Intra Cos: 0.9994824528694153
Inter Cos: 0.11822619289159775
Norm Quadratic Average: 8.923516273498535
Nearest Class Center Accuracy: 0.9999666666666667

Test Set:
Average Loss: 0.016786879842728378
Accuracy: 0.996
NC1 Within Class Collapse: 0.07430818676948547
NC2 Equinorm: Features: 0.01486364658921957, Weights: 0.005436608102172613
NC2 Equiangle: Features: 0.09996077219645182, Weights: 0.0627908123864068
NC3 Self-Duality: 0.025389457121491432
NC4 NCC Mismatch: 9.999999999998899e-05

Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10219582915306091
Inter Cos: 0.12048853933811188
Norm Quadratic Average: 23.595195770263672
Nearest Class Center Accuracy: 0.8229

Layer 1: Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.0674624890089035
Inter Cos: 0.0798259973526001
Norm Quadratic Average: 2.508967876434326
Nearest Class Center Accuracy: 0.8214

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.11404064297676086
Inter Cos: 0.10592646151781082
Norm Quadratic Average: 1.5399935245513916
Nearest Class Center Accuracy: 0.8842

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10600379854440689
Inter Cos: 0.10140937566757202
Norm Quadratic Average: 1.2004235982894897
Nearest Class Center Accuracy: 0.8883

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.18792355060577393
Inter Cos: 0.12298475950956345
Norm Quadratic Average: 0.821699321269989
Nearest Class Center Accuracy: 0.943

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.24451634287834167
Inter Cos: 0.13089551031589508
Norm Quadratic Average: 0.6089138984680176
Nearest Class Center Accuracy: 0.9639

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.31986716389656067
Inter Cos: 0.14349471032619476
Norm Quadratic Average: 0.5115757584571838
Nearest Class Center Accuracy: 0.9738

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3654223382472992
Inter Cos: 0.14616858959197998
Norm Quadratic Average: 0.4425000846385956
Nearest Class Center Accuracy: 0.9775

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4335871636867523
Inter Cos: 0.16661220788955688
Norm Quadratic Average: 0.28511956334114075
Nearest Class Center Accuracy: 0.989

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6638658046722412
Inter Cos: 0.27126041054725647
Norm Quadratic Average: 0.1912594586610794
Nearest Class Center Accuracy: 0.9942

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8045987486839294
Inter Cos: 0.2721133232116699
Norm Quadratic Average: 0.19180911779403687
Nearest Class Center Accuracy: 0.9955

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8461460471153259
Inter Cos: 0.1671903431415558
Norm Quadratic Average: 0.18813163042068481
Nearest Class Center Accuracy: 0.9952

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9036628603935242
Inter Cos: 0.17757391929626465
Norm Quadratic Average: 0.22692008316516876
Nearest Class Center Accuracy: 0.9958

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9767321944236755
Inter Cos: 0.040296345949172974
Norm Quadratic Average: 0.2724522054195404
Nearest Class Center Accuracy: 0.9961

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9800291061401367
Inter Cos: 0.005810700356960297
Norm Quadratic Average: 0.532344400882721
Nearest Class Center Accuracy: 0.9961

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9811743497848511
Inter Cos: 0.04251845180988312
Norm Quadratic Average: 1.0999764204025269
Nearest Class Center Accuracy: 0.9961

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.144974708557129
Linear Weight Rank: 10
Intra Cos: 0.9818337559700012
Inter Cos: 0.13037315011024475
Norm Quadratic Average: 24.974605560302734
Nearest Class Center Accuracy: 0.9961

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.1474530696868896
Linear Weight Rank: 1426
Intra Cos: 0.9827648997306824
Inter Cos: 0.18210764229297638
Norm Quadratic Average: 17.14148712158203
Nearest Class Center Accuracy: 0.9961

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.1475274562835693
Linear Weight Rank: 9
Intra Cos: 0.982906699180603
Inter Cos: 0.1607092171907425
Norm Quadratic Average: 11.998644828796387
Nearest Class Center Accuracy: 0.9961

Output Layer:
Intra Cos: 0.9837188124656677
Inter Cos: 0.11751573532819748
Norm Quadratic Average: 8.888669967651367
Nearest Class Center Accuracy: 0.9961

