Model save path: ./New_Models/bn_False_dataset_MNIST_epochs_200_lr_0.01_model_type_vgg19_rand_seed_338327_test_samples_None_train_samples_None_weight_decay_0.001.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: 0.10967152565717697
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: 0.09927153587341309
Inter Cos: 0.1203032061457634
Norm Quadratic Average: 59.43273162841797
Nearest Class Center Accuracy: 0.8081

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.12526485323905945
Inter Cos: 0.1504223495721817
Norm Quadratic Average: 72.40592193603516
Nearest Class Center Accuracy: 0.81555

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1338283270597458
Inter Cos: 0.16280469298362732
Norm Quadratic Average: 111.12981414794922
Nearest Class Center Accuracy: 0.82365

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2004457414150238
Inter Cos: 0.18225257098674774
Norm Quadratic Average: 83.05243682861328
Nearest Class Center Accuracy: 0.8795333333333333

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.23342245817184448
Inter Cos: 0.1858287900686264
Norm Quadratic Average: 71.9631118774414
Nearest Class Center Accuracy: 0.9111666666666667

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2677287757396698
Inter Cos: 0.19686731696128845
Norm Quadratic Average: 56.13462829589844
Nearest Class Center Accuracy: 0.9379666666666666

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2995207905769348
Inter Cos: 0.19334962964057922
Norm Quadratic Average: 37.47564697265625
Nearest Class Center Accuracy: 0.9554833333333334

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.33461517095565796
Inter Cos: 0.20735305547714233
Norm Quadratic Average: 14.11640453338623
Nearest Class Center Accuracy: 0.9731166666666666

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5355318188667297
Inter Cos: 0.30089715123176575
Norm Quadratic Average: 8.51402759552002
Nearest Class Center Accuracy: 0.98095

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6254205107688904
Inter Cos: 0.33607953786849976
Norm Quadratic Average: 9.612370491027832
Nearest Class Center Accuracy: 0.9827833333333333

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6903618574142456
Inter Cos: 0.3348635137081146
Norm Quadratic Average: 12.659917831420898
Nearest Class Center Accuracy: 0.9871166666666666

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7154428362846375
Inter Cos: 0.26506224274635315
Norm Quadratic Average: 9.830050468444824
Nearest Class Center Accuracy: 0.9801333333333333

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8253552913665771
Inter Cos: 0.32786059379577637
Norm Quadratic Average: 9.005897521972656
Nearest Class Center Accuracy: 0.9863333333333333

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8580194115638733
Inter Cos: 0.34456831216812134
Norm Quadratic Average: 9.961872100830078
Nearest Class Center Accuracy: 0.9906666666666667

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8759335279464722
Inter Cos: 0.3476782441139221
Norm Quadratic Average: 10.40809154510498
Nearest Class Center Accuracy: 0.9932666666666666

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.884976625442505
Linear Weight Rank: 4029
Intra Cos: 0.8914250731468201
Inter Cos: 0.3313556909561157
Norm Quadratic Average: 49.93428421020508
Nearest Class Center Accuracy: 0.9949666666666667

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.26589298248291
Linear Weight Rank: 3640
Intra Cos: 0.9194202423095703
Inter Cos: 0.32017627358436584
Norm Quadratic Average: 39.08393096923828
Nearest Class Center Accuracy: 0.99725

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.019822597503662
Linear Weight Rank: 9
Intra Cos: 0.9256513714790344
Inter Cos: 0.2883600890636444
Norm Quadratic Average: 28.666040420532227
Nearest Class Center Accuracy: 0.9986166666666667

Output Layer:
Intra Cos: 0.9529868960380554
Inter Cos: 0.33169612288475037
Norm Quadratic Average: 23.14618492126465
Nearest Class Center Accuracy: 0.9997666666666667

Test Set:
Average Loss: 0.022906164551966868
Accuracy: 0.9935
NC1 Within Class Collapse: 0.8642585277557373
NC2 Equinorm: Features: 0.10521373152732849, Weights: 0.03491558879613876
NC2 Equiangle: Features: 0.2696272956000434, Weights: 0.18389305538601344
NC3 Self-Duality: 0.1192842647433281
NC4 NCC Mismatch: 0.0042999999999999705

Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10219582170248032
Inter Cos: 0.12048853188753128
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: 0.11070387810468674
Inter Cos: 0.13233134150505066
Norm Quadratic Average: 59.625083923339844
Nearest Class Center Accuracy: 0.8206

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.13948290050029755
Inter Cos: 0.1646861433982849
Norm Quadratic Average: 72.44918060302734
Nearest Class Center Accuracy: 0.83

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1479061096906662
Inter Cos: 0.17808854579925537
Norm Quadratic Average: 111.26388549804688
Nearest Class Center Accuracy: 0.8366

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.216229647397995
Inter Cos: 0.19875270128250122
Norm Quadratic Average: 83.07052612304688
Nearest Class Center Accuracy: 0.8889

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.25004804134368896
Inter Cos: 0.20086176693439484
Norm Quadratic Average: 72.0241470336914
Nearest Class Center Accuracy: 0.9221

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.283469021320343
Inter Cos: 0.20699575543403625
Norm Quadratic Average: 56.263145446777344
Nearest Class Center Accuracy: 0.9478

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3138667047023773
Inter Cos: 0.20920342206954956
Norm Quadratic Average: 37.605690002441406
Nearest Class Center Accuracy: 0.9614

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.35015732049942017
Inter Cos: 0.23045657575130463
Norm Quadratic Average: 14.173550605773926
Nearest Class Center Accuracy: 0.9743

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5576017498970032
Inter Cos: 0.31727907061576843
Norm Quadratic Average: 8.571388244628906
Nearest Class Center Accuracy: 0.9761

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6461828947067261
Inter Cos: 0.3519085645675659
Norm Quadratic Average: 9.694670677185059
Nearest Class Center Accuracy: 0.9775

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7087920904159546
Inter Cos: 0.36417657136917114
Norm Quadratic Average: 12.7833833694458
Nearest Class Center Accuracy: 0.982

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7201809287071228
Inter Cos: 0.2739134728908539
Norm Quadratic Average: 9.939343452453613
Nearest Class Center Accuracy: 0.975

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8361145853996277
Inter Cos: 0.35120368003845215
Norm Quadratic Average: 9.11475658416748
Nearest Class Center Accuracy: 0.9786

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8669748902320862
Inter Cos: 0.34583377838134766
Norm Quadratic Average: 10.080655097961426
Nearest Class Center Accuracy: 0.9825

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8816233277320862
Inter Cos: 0.3505324423313141
Norm Quadratic Average: 10.525433540344238
Nearest Class Center Accuracy: 0.9851

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.884976625442505
Linear Weight Rank: 4029
Intra Cos: 0.895855724811554
Inter Cos: 0.32366010546684265
Norm Quadratic Average: 50.38066482543945
Nearest Class Center Accuracy: 0.9877

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.26589298248291
Linear Weight Rank: 3640
Intra Cos: 0.9164023399353027
Inter Cos: 0.31392577290534973
Norm Quadratic Average: 39.41020584106445
Nearest Class Center Accuracy: 0.9908

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.019822597503662
Linear Weight Rank: 9
Intra Cos: 0.9213849902153015
Inter Cos: 0.28265324234962463
Norm Quadratic Average: 28.901077270507812
Nearest Class Center Accuracy: 0.9917

Output Layer:
Intra Cos: 0.9449811577796936
Inter Cos: 0.3300054967403412
Norm Quadratic Average: 23.327560424804688
Nearest Class Center Accuracy: 0.9922

