Model save path: ./New_Models/bn_False_dataset_MNIST_epochs_200_lr_0.01_model_type_vgg19_rand_seed_979323_test_samples_None_train_samples_None_weight_decay_0.0001.pth.tar
Training Set:
Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.09116754680871964
Inter Cos: 0.10967149585485458
Norm Quadratic Average: 23.567672729492188
Nearest Class Center Accuracy: 0.8079833333333334

Layer 1: Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.08298904448747635
Inter Cos: 0.09935929626226425
Norm Quadratic Average: 55.83138656616211
Nearest Class Center Accuracy: 0.81835

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.12394872307777405
Inter Cos: 0.13100898265838623
Norm Quadratic Average: 51.01303482055664
Nearest Class Center Accuracy: 0.853

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1328517645597458
Inter Cos: 0.13904288411140442
Norm Quadratic Average: 70.34307861328125
Nearest Class Center Accuracy: 0.8599

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2102614790201187
Inter Cos: 0.16753144562244415
Norm Quadratic Average: 50.882102966308594
Nearest Class Center Accuracy: 0.917

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.23605357110500336
Inter Cos: 0.16932685673236847
Norm Quadratic Average: 49.24711227416992
Nearest Class Center Accuracy: 0.9408666666666666

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.26171377301216125
Inter Cos: 0.16726361215114594
Norm Quadratic Average: 44.34465408325195
Nearest Class Center Accuracy: 0.9546833333333333

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.29471009969711304
Inter Cos: 0.17469459772109985
Norm Quadratic Average: 40.26394271850586
Nearest Class Center Accuracy: 0.9623833333333334

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.34836986660957336
Inter Cos: 0.1564873307943344
Norm Quadratic Average: 20.221406936645508
Nearest Class Center Accuracy: 0.9835166666666667

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.47067388892173767
Inter Cos: 0.16945354640483856
Norm Quadratic Average: 16.973527908325195
Nearest Class Center Accuracy: 0.991

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5839154124259949
Inter Cos: 0.19802501797676086
Norm Quadratic Average: 15.822128295898438
Nearest Class Center Accuracy: 0.99455

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6637502312660217
Inter Cos: 0.20684529840946198
Norm Quadratic Average: 15.136384963989258
Nearest Class Center Accuracy: 0.9967

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7199372053146362
Inter Cos: 0.16595806181430817
Norm Quadratic Average: 9.731571197509766
Nearest Class Center Accuracy: 0.9976833333333334

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8696329593658447
Inter Cos: 0.1853458434343338
Norm Quadratic Average: 8.69005298614502
Nearest Class Center Accuracy: 0.9986833333333334

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9277645945549011
Inter Cos: 0.26903221011161804
Norm Quadratic Average: 8.05408763885498
Nearest Class Center Accuracy: 0.99875

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9458931684494019
Inter Cos: 0.308794230222702
Norm Quadratic Average: 7.430237770080566
Nearest Class Center Accuracy: 0.99905

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 68.78544616699219
Linear Weight Rank: 4031
Intra Cos: 0.9571194648742676
Inter Cos: 0.2813001573085785
Norm Quadratic Average: 41.32009506225586
Nearest Class Center Accuracy: 0.9995333333333334

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 27.895915985107422
Linear Weight Rank: 3670
Intra Cos: 0.9582562446594238
Inter Cos: 0.27106258273124695
Norm Quadratic Average: 32.79326248168945
Nearest Class Center Accuracy: 0.9997666666666667

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.808513879776001
Linear Weight Rank: 10
Intra Cos: 0.9572946429252625
Inter Cos: 0.21862801909446716
Norm Quadratic Average: 27.899282455444336
Nearest Class Center Accuracy: 0.9999166666666667

Output Layer:
Intra Cos: 0.9867953658103943
Inter Cos: 0.26326414942741394
Norm Quadratic Average: 25.6868896484375
Nearest Class Center Accuracy: 1.0

Test Set:
Average Loss: 0.026740946345473002
Accuracy: 0.9946
NC1 Within Class Collapse: 0.4088476300239563
NC2 Equinorm: Features: 0.08310043811798096, Weights: 0.03657413646578789
NC2 Equiangle: Features: 0.19415465460883247, Weights: 0.10868879954020182
NC3 Self-Duality: 0.28418946266174316
NC4 NCC Mismatch: 0.0027000000000000357

Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10219583660364151
Inter Cos: 0.12048852443695068
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: 0.09332451969385147
Inter Cos: 0.10714765638113022
Norm Quadratic Average: 55.83274841308594
Nearest Class Center Accuracy: 0.831

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1368214339017868
Inter Cos: 0.14005056023597717
Norm Quadratic Average: 50.83183670043945
Nearest Class Center Accuracy: 0.8658

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.14538346230983734
Inter Cos: 0.14996932446956635
Norm Quadratic Average: 70.20647430419922
Nearest Class Center Accuracy: 0.8711

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2277528941631317
Inter Cos: 0.18157492578029633
Norm Quadratic Average: 50.772743225097656
Nearest Class Center Accuracy: 0.9263

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.252616286277771
Inter Cos: 0.18431203067302704
Norm Quadratic Average: 49.20158767700195
Nearest Class Center Accuracy: 0.9477

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.278315007686615
Inter Cos: 0.18252046406269073
Norm Quadratic Average: 44.31894302368164
Nearest Class Center Accuracy: 0.9587

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.31091177463531494
Inter Cos: 0.19073092937469482
Norm Quadratic Average: 40.243709564208984
Nearest Class Center Accuracy: 0.9646

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3603781759738922
Inter Cos: 0.1705918163061142
Norm Quadratic Average: 20.231210708618164
Nearest Class Center Accuracy: 0.9832

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.483523428440094
Inter Cos: 0.18547222018241882
Norm Quadratic Average: 17.008237838745117
Nearest Class Center Accuracy: 0.9891

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5950188636779785
Inter Cos: 0.21418626606464386
Norm Quadratic Average: 15.871182441711426
Nearest Class Center Accuracy: 0.9907

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6713308095932007
Inter Cos: 0.2218651920557022
Norm Quadratic Average: 15.196928024291992
Nearest Class Center Accuracy: 0.9909

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7216142416000366
Inter Cos: 0.1641939878463745
Norm Quadratic Average: 9.77526569366455
Nearest Class Center Accuracy: 0.9918

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8672894835472107
Inter Cos: 0.1862488090991974
Norm Quadratic Average: 8.733518600463867
Nearest Class Center Accuracy: 0.9924

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9233875870704651
Inter Cos: 0.268584668636322
Norm Quadratic Average: 8.092934608459473
Nearest Class Center Accuracy: 0.9927

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9405869245529175
Inter Cos: 0.3070155084133148
Norm Quadratic Average: 7.463862895965576
Nearest Class Center Accuracy: 0.9931

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 68.78544616699219
Linear Weight Rank: 4031
Intra Cos: 0.9492450952529907
Inter Cos: 0.27856093645095825
Norm Quadratic Average: 41.50075149536133
Nearest Class Center Accuracy: 0.9936

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 27.895915985107422
Linear Weight Rank: 3670
Intra Cos: 0.9504210948944092
Inter Cos: 0.2677934765815735
Norm Quadratic Average: 32.94024658203125
Nearest Class Center Accuracy: 0.994

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.808513879776001
Linear Weight Rank: 10
Intra Cos: 0.9489009976387024
Inter Cos: 0.21601933240890503
Norm Quadratic Average: 28.023714065551758
Nearest Class Center Accuracy: 0.9942

Output Layer:
Intra Cos: 0.974653422832489
Inter Cos: 0.2619199752807617
Norm Quadratic Average: 25.798954010009766
Nearest Class Center Accuracy: 0.9948

