Model save path: ./New_Models/bn_True_dataset_MNIST_epochs_200_lr_0.01_model_type_vgg19_rand_seed_846264_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: 0.10967151820659637
Norm Quadratic Average: 23.567686080932617
Nearest Class Center Accuracy: 0.8079833333333334

Layer 1: Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.06663807481527328
Inter Cos: 0.07903498411178589
Norm Quadratic Average: 99.2023696899414
Nearest Class Center Accuracy: 0.8266166666666667

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10677798092365265
Inter Cos: 0.1046384647488594
Norm Quadratic Average: 62.67282485961914
Nearest Class Center Accuracy: 0.8653833333333333

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.11230861395597458
Inter Cos: 0.11118323355913162
Norm Quadratic Average: 63.23434829711914
Nearest Class Center Accuracy: 0.8765333333333334

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.18770042061805725
Inter Cos: 0.13502156734466553
Norm Quadratic Average: 39.9150276184082
Nearest Class Center Accuracy: 0.9263333333333333

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.22001810371875763
Inter Cos: 0.1389589011669159
Norm Quadratic Average: 40.56741714477539
Nearest Class Center Accuracy: 0.946

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.242304727435112
Inter Cos: 0.13842761516571045
Norm Quadratic Average: 42.28767395019531
Nearest Class Center Accuracy: 0.9603833333333334

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.26413819193840027
Inter Cos: 0.1315891295671463
Norm Quadratic Average: 42.816566467285156
Nearest Class Center Accuracy: 0.9693

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3158232271671295
Inter Cos: 0.11373575776815414
Norm Quadratic Average: 28.60562515258789
Nearest Class Center Accuracy: 0.9892833333333333

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3889971971511841
Inter Cos: 0.13505695760250092
Norm Quadratic Average: 29.074636459350586
Nearest Class Center Accuracy: 0.9950833333333333

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4885990023612976
Inter Cos: 0.13669085502624512
Norm Quadratic Average: 30.3256893157959
Nearest Class Center Accuracy: 0.9982

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5817343592643738
Inter Cos: 0.140994593501091
Norm Quadratic Average: 30.927114486694336
Nearest Class Center Accuracy: 0.9993666666666666

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7003053426742554
Inter Cos: 0.20336975157260895
Norm Quadratic Average: 24.345186233520508
Nearest Class Center Accuracy: 0.99985

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8457898497581482
Inter Cos: 0.18252822756767273
Norm Quadratic Average: 16.22706413269043
Nearest Class Center Accuracy: 0.9999666666666667

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9018852710723877
Inter Cos: 0.17000319063663483
Norm Quadratic Average: 17.30654525756836
Nearest Class Center Accuracy: 1.0

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9327142238616943
Inter Cos: 0.08308123797178268
Norm Quadratic Average: 18.23462677001953
Nearest Class Center Accuracy: 1.0

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 68.81282806396484
Linear Weight Rank: 4031
Intra Cos: 0.9527614712715149
Inter Cos: 0.06562215089797974
Norm Quadratic Average: 122.23085021972656
Nearest Class Center Accuracy: 1.0

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 27.938892364501953
Linear Weight Rank: 3670
Intra Cos: 0.9856282472610474
Inter Cos: 0.0911816656589508
Norm Quadratic Average: 64.07052612304688
Nearest Class Center Accuracy: 1.0

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.390734910964966
Linear Weight Rank: 10
Intra Cos: 0.9860844016075134
Inter Cos: 0.10172697901725769
Norm Quadratic Average: 30.871654510498047
Nearest Class Center Accuracy: 1.0

Output Layer:
Intra Cos: 0.9982805848121643
Inter Cos: 0.19277441501617432
Norm Quadratic Average: 18.56580924987793
Nearest Class Center Accuracy: 1.0

Test Set:
Average Loss: 0.023146734803839355
Accuracy: 0.9953
NC1 Within Class Collapse: 0.19868415594100952
NC2 Equinorm: Features: 0.05225944146513939, Weights: 0.013932651840150356
NC2 Equiangle: Features: 0.08552244504292807, Weights: 0.08142325083414713
NC3 Self-Duality: 0.5758136510848999
NC4 NCC Mismatch: 0.0007000000000000339

Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10219582170248032
Inter Cos: 0.12048851698637009
Norm Quadratic Average: 23.59519386291504
Nearest Class Center Accuracy: 0.8229

Layer 1: Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.07488849014043808
Inter Cos: 0.08235624432563782
Norm Quadratic Average: 98.96360778808594
Nearest Class Center Accuracy: 0.8401

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.11844682693481445
Inter Cos: 0.10491662472486496
Norm Quadratic Average: 62.25550842285156
Nearest Class Center Accuracy: 0.8765

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1239558532834053
Inter Cos: 0.1146715059876442
Norm Quadratic Average: 62.78964614868164
Nearest Class Center Accuracy: 0.8858

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.20321986079216003
Inter Cos: 0.14790993928909302
Norm Quadratic Average: 39.61979293823242
Nearest Class Center Accuracy: 0.9365

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.23633241653442383
Inter Cos: 0.15249967575073242
Norm Quadratic Average: 40.280696868896484
Nearest Class Center Accuracy: 0.9524

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.25595593452453613
Inter Cos: 0.15331515669822693
Norm Quadratic Average: 42.03022766113281
Nearest Class Center Accuracy: 0.9643

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.27682632207870483
Inter Cos: 0.1465044468641281
Norm Quadratic Average: 42.60983657836914
Nearest Class Center Accuracy: 0.972

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.32754838466644287
Inter Cos: 0.11586498469114304
Norm Quadratic Average: 28.509178161621094
Nearest Class Center Accuracy: 0.9886

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.39757296442985535
Inter Cos: 0.13456718623638153
Norm Quadratic Average: 29.021562576293945
Nearest Class Center Accuracy: 0.9913

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.49425208568573
Inter Cos: 0.14058557152748108
Norm Quadratic Average: 30.301097869873047
Nearest Class Center Accuracy: 0.9925

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5843477845191956
Inter Cos: 0.1358523666858673
Norm Quadratic Average: 30.922603607177734
Nearest Class Center Accuracy: 0.9932

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6987975835800171
Inter Cos: 0.1941157728433609
Norm Quadratic Average: 24.356821060180664
Nearest Class Center Accuracy: 0.9935

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8396990299224854
Inter Cos: 0.17252829670906067
Norm Quadratic Average: 16.24123191833496
Nearest Class Center Accuracy: 0.994

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8957056999206543
Inter Cos: 0.1603323519229889
Norm Quadratic Average: 17.323328018188477
Nearest Class Center Accuracy: 0.9946

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9226040244102478
Inter Cos: 0.0745563730597496
Norm Quadratic Average: 18.250288009643555
Nearest Class Center Accuracy: 0.9949

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 68.81282806396484
Linear Weight Rank: 4031
Intra Cos: 0.9269360899925232
Inter Cos: 0.06978904455900192
Norm Quadratic Average: 122.35224151611328
Nearest Class Center Accuracy: 0.9947

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 27.938892364501953
Linear Weight Rank: 3670
Intra Cos: 0.9599411487579346
Inter Cos: 0.0937914103269577
Norm Quadratic Average: 64.04997253417969
Nearest Class Center Accuracy: 0.9953

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.390734910964966
Linear Weight Rank: 10
Intra Cos: 0.9588747024536133
Inter Cos: 0.10342549532651901
Norm Quadratic Average: 30.864089965820312
Nearest Class Center Accuracy: 0.9954

Output Layer:
Intra Cos: 0.9844266176223755
Inter Cos: 0.2006634771823883
Norm Quadratic Average: 18.53810691833496
Nearest Class Center Accuracy: 0.995

