Model save path: ./New_Models/bn_False_dataset_MNIST_epochs_200_lr_0.01_model_type_vgg19_rand_seed_846264_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: 0.09116754680871964
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: 0.09079409390687943
Inter Cos: 0.10920485854148865
Norm Quadratic Average: 55.333499908447266
Nearest Class Center Accuracy: 0.8128166666666666

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.12447302788496017
Inter Cos: 0.13481749594211578
Norm Quadratic Average: 56.04275894165039
Nearest Class Center Accuracy: 0.84155

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.13848944008350372
Inter Cos: 0.15164726972579956
Norm Quadratic Average: 78.95716857910156
Nearest Class Center Accuracy: 0.8493

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.21944105625152588
Inter Cos: 0.17895357310771942
Norm Quadratic Average: 63.24003982543945
Nearest Class Center Accuracy: 0.8977

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.24495349824428558
Inter Cos: 0.18919141590595245
Norm Quadratic Average: 63.22236251831055
Nearest Class Center Accuracy: 0.9189833333333334

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.26544269919395447
Inter Cos: 0.1934226155281067
Norm Quadratic Average: 56.57978439331055
Nearest Class Center Accuracy: 0.9351333333333334

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.28802385926246643
Inter Cos: 0.1812218278646469
Norm Quadratic Average: 43.36327362060547
Nearest Class Center Accuracy: 0.9511166666666667

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3310752809047699
Inter Cos: 0.16052395105361938
Norm Quadratic Average: 18.5837345123291
Nearest Class Center Accuracy: 0.9713666666666667

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.49409618973731995
Inter Cos: 0.26340216398239136
Norm Quadratic Average: 11.707053184509277
Nearest Class Center Accuracy: 0.9846666666666667

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5868546962738037
Inter Cos: 0.30584314465522766
Norm Quadratic Average: 10.733020782470703
Nearest Class Center Accuracy: 0.98885

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6418118476867676
Inter Cos: 0.29991185665130615
Norm Quadratic Average: 10.91675853729248
Nearest Class Center Accuracy: 0.9925666666666667

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.698486864566803
Inter Cos: 0.28257641196250916
Norm Quadratic Average: 6.860213279724121
Nearest Class Center Accuracy: 0.9915166666666667

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8570525050163269
Inter Cos: 0.30983585119247437
Norm Quadratic Average: 6.0075602531433105
Nearest Class Center Accuracy: 0.99405

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8921775221824646
Inter Cos: 0.2999344766139984
Norm Quadratic Average: 6.184818744659424
Nearest Class Center Accuracy: 0.99485

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.910178005695343
Inter Cos: 0.3429570496082306
Norm Quadratic Average: 6.0781474113464355
Nearest Class Center Accuracy: 0.9956333333333334

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 14.710909843444824
Linear Weight Rank: 4031
Intra Cos: 0.9172877073287964
Inter Cos: 0.30829155445098877
Norm Quadratic Average: 34.13840103149414
Nearest Class Center Accuracy: 0.9969

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 6.350213050842285
Linear Weight Rank: 3667
Intra Cos: 0.9312398433685303
Inter Cos: 0.29508230090141296
Norm Quadratic Average: 30.094697952270508
Nearest Class Center Accuracy: 0.9978833333333333

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.3573386669158936
Linear Weight Rank: 10
Intra Cos: 0.938416600227356
Inter Cos: 0.2566910982131958
Norm Quadratic Average: 26.066898345947266
Nearest Class Center Accuracy: 0.9985833333333334

Output Layer:
Intra Cos: 0.9649690985679626
Inter Cos: 0.38431066274642944
Norm Quadratic Average: 25.099611282348633
Nearest Class Center Accuracy: 0.9997333333333334

Test Set:
Average Loss: 0.022726654061954468
Accuracy: 0.9948
NC1 Within Class Collapse: 0.9831647872924805
NC2 Equinorm: Features: 0.1443335860967636, Weights: 0.024963967502117157
NC2 Equiangle: Features: 0.2582184261745877, Weights: 0.19255958133273654
NC3 Self-Duality: 0.14982998371124268
NC4 NCC Mismatch: 0.0050000000000000044

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: 0.10164964199066162
Inter Cos: 0.12004949152469635
Norm Quadratic Average: 55.43150329589844
Nearest Class Center Accuracy: 0.8246

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1380205601453781
Inter Cos: 0.14570200443267822
Norm Quadratic Average: 55.96055221557617
Nearest Class Center Accuracy: 0.8548

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1525517702102661
Inter Cos: 0.1655847430229187
Norm Quadratic Average: 78.91943359375
Nearest Class Center Accuracy: 0.8606

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.23849734663963318
Inter Cos: 0.19408652186393738
Norm Quadratic Average: 63.162776947021484
Nearest Class Center Accuracy: 0.9088

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.26513004302978516
Inter Cos: 0.19299370050430298
Norm Quadratic Average: 63.12593460083008
Nearest Class Center Accuracy: 0.9302

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2836240530014038
Inter Cos: 0.19312041997909546
Norm Quadratic Average: 56.50730895996094
Nearest Class Center Accuracy: 0.9441

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.30483442544937134
Inter Cos: 0.19756051898002625
Norm Quadratic Average: 43.37433624267578
Nearest Class Center Accuracy: 0.9575

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.34812676906585693
Inter Cos: 0.17165662348270416
Norm Quadratic Average: 18.613601684570312
Nearest Class Center Accuracy: 0.9736

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5199527144432068
Inter Cos: 0.2821296155452728
Norm Quadratic Average: 11.757944107055664
Nearest Class Center Accuracy: 0.9815

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6133697628974915
Inter Cos: 0.3266906142234802
Norm Quadratic Average: 10.79669189453125
Nearest Class Center Accuracy: 0.9835

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6649855375289917
Inter Cos: 0.3202161490917206
Norm Quadratic Average: 10.9881010055542
Nearest Class Center Accuracy: 0.9878

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7029911279678345
Inter Cos: 0.2873992621898651
Norm Quadratic Average: 6.913942337036133
Nearest Class Center Accuracy: 0.985

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8692638874053955
Inter Cos: 0.3139002323150635
Norm Quadratic Average: 6.065098285675049
Nearest Class Center Accuracy: 0.9877

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8981389403343201
Inter Cos: 0.2950887680053711
Norm Quadratic Average: 6.241555213928223
Nearest Class Center Accuracy: 0.9889

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9075050950050354
Inter Cos: 0.33606189489364624
Norm Quadratic Average: 6.130014419555664
Nearest Class Center Accuracy: 0.9894

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 14.710909843444824
Linear Weight Rank: 4031
Intra Cos: 0.9131323099136353
Inter Cos: 0.3056233823299408
Norm Quadratic Average: 34.413063049316406
Nearest Class Center Accuracy: 0.9905

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 6.350213050842285
Linear Weight Rank: 3667
Intra Cos: 0.9263314008712769
Inter Cos: 0.2925829589366913
Norm Quadratic Average: 30.32813835144043
Nearest Class Center Accuracy: 0.9912

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.3573386669158936
Linear Weight Rank: 10
Intra Cos: 0.932331919670105
Inter Cos: 0.2561382055282593
Norm Quadratic Average: 26.261497497558594
Nearest Class Center Accuracy: 0.9917

Output Layer:
Intra Cos: 0.9570133686065674
Inter Cos: 0.3833841383457184
Norm Quadratic Average: 25.270626068115234
Nearest Class Center Accuracy: 0.9941

