Model save path: ./New_Models/bn_True_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: 0.06426165252923965
Inter Cos: 0.08006240427494049
Norm Quadratic Average: 93.84632873535156
Nearest Class Center Accuracy: 0.8249666666666666

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1015666052699089
Inter Cos: 0.10571162402629852
Norm Quadratic Average: 57.22101974487305
Nearest Class Center Accuracy: 0.8634333333333334

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.11158943176269531
Inter Cos: 0.10858609527349472
Norm Quadratic Average: 62.659019470214844
Nearest Class Center Accuracy: 0.8816666666666667

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1801392287015915
Inter Cos: 0.14289340376853943
Norm Quadratic Average: 38.94818115234375
Nearest Class Center Accuracy: 0.9295666666666667

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2114194631576538
Inter Cos: 0.14534054696559906
Norm Quadratic Average: 39.50718307495117
Nearest Class Center Accuracy: 0.94925

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.23584076762199402
Inter Cos: 0.1466083824634552
Norm Quadratic Average: 42.240684509277344
Nearest Class Center Accuracy: 0.9637333333333333

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2615067958831787
Inter Cos: 0.13547779619693756
Norm Quadratic Average: 43.09503936767578
Nearest Class Center Accuracy: 0.9729666666666666

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.32294753193855286
Inter Cos: 0.11660676449537277
Norm Quadratic Average: 28.590557098388672
Nearest Class Center Accuracy: 0.9912333333333333

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4130364656448364
Inter Cos: 0.1635211706161499
Norm Quadratic Average: 29.64129066467285
Nearest Class Center Accuracy: 0.9967

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5104683637619019
Inter Cos: 0.18462993204593658
Norm Quadratic Average: 30.855424880981445
Nearest Class Center Accuracy: 0.9986333333333334

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5967110395431519
Inter Cos: 0.16927482187747955
Norm Quadratic Average: 31.23604393005371
Nearest Class Center Accuracy: 0.9994333333333333

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7113426923751831
Inter Cos: 0.27666807174682617
Norm Quadratic Average: 25.043235778808594
Nearest Class Center Accuracy: 0.9995

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8417156934738159
Inter Cos: 0.17873549461364746
Norm Quadratic Average: 16.63600730895996
Nearest Class Center Accuracy: 0.9999166666666667

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8994807600975037
Inter Cos: 0.13990464806556702
Norm Quadratic Average: 17.436464309692383
Nearest Class Center Accuracy: 1.0

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9339931607246399
Inter Cos: 0.11790478229522705
Norm Quadratic Average: 18.39632797241211
Nearest Class Center Accuracy: 1.0

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 68.80902862548828
Linear Weight Rank: 4031
Intra Cos: 0.952033519744873
Inter Cos: 0.10436870157718658
Norm Quadratic Average: 122.91443634033203
Nearest Class Center Accuracy: 1.0

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 27.94831085205078
Linear Weight Rank: 3670
Intra Cos: 0.9860516786575317
Inter Cos: 0.05809307098388672
Norm Quadratic Average: 64.32003784179688
Nearest Class Center Accuracy: 1.0

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.4492595195770264
Linear Weight Rank: 10
Intra Cos: 0.986225962638855
Inter Cos: 0.057255443185567856
Norm Quadratic Average: 31.12812614440918
Nearest Class Center Accuracy: 1.0

Output Layer:
Intra Cos: 0.9984179735183716
Inter Cos: 0.30745381116867065
Norm Quadratic Average: 18.874103546142578
Nearest Class Center Accuracy: 1.0

Test Set:
Average Loss: 0.024501776975263966
Accuracy: 0.9958
NC1 Within Class Collapse: 0.2087327092885971
NC2 Equinorm: Features: 0.0596604198217392, Weights: 0.02847759798169136
NC2 Equiangle: Features: 0.07875921461317274, Weights: 0.08009365929497612
NC3 Self-Duality: 0.5922362208366394
NC4 NCC Mismatch: 0.00039999999999995595

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: 0.0725022405385971
Inter Cos: 0.08117613196372986
Norm Quadratic Average: 93.57623291015625
Nearest Class Center Accuracy: 0.8366

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.11244846880435944
Inter Cos: 0.10645865648984909
Norm Quadratic Average: 56.847530364990234
Nearest Class Center Accuracy: 0.8756

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.12305592000484467
Inter Cos: 0.11728031933307648
Norm Quadratic Average: 62.251033782958984
Nearest Class Center Accuracy: 0.8938

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.19589127600193024
Inter Cos: 0.15721988677978516
Norm Quadratic Average: 38.68947219848633
Nearest Class Center Accuracy: 0.9386

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.22335213422775269
Inter Cos: 0.1598641574382782
Norm Quadratic Average: 39.269351959228516
Nearest Class Center Accuracy: 0.9545

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2476881444454193
Inter Cos: 0.16104768216609955
Norm Quadratic Average: 42.04157257080078
Nearest Class Center Accuracy: 0.9654

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2745886445045471
Inter Cos: 0.15012182295322418
Norm Quadratic Average: 42.93716049194336
Nearest Class Center Accuracy: 0.9733

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.33625710010528564
Inter Cos: 0.12986840307712555
Norm Quadratic Average: 28.524066925048828
Nearest Class Center Accuracy: 0.9886

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4248313307762146
Inter Cos: 0.16822780668735504
Norm Quadratic Average: 29.593902587890625
Nearest Class Center Accuracy: 0.9913

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5181277394294739
Inter Cos: 0.18835192918777466
Norm Quadratic Average: 30.820850372314453
Nearest Class Center Accuracy: 0.9934

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6006720662117004
Inter Cos: 0.1704089343547821
Norm Quadratic Average: 31.20530891418457
Nearest Class Center Accuracy: 0.9942

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.706271767616272
Inter Cos: 0.27465757727622986
Norm Quadratic Average: 25.025535583496094
Nearest Class Center Accuracy: 0.9931

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8298302888870239
Inter Cos: 0.17654374241828918
Norm Quadratic Average: 16.622318267822266
Nearest Class Center Accuracy: 0.994

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8848611116409302
Inter Cos: 0.13822093605995178
Norm Quadratic Average: 17.42338752746582
Nearest Class Center Accuracy: 0.9946

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.918615460395813
Inter Cos: 0.11755302548408508
Norm Quadratic Average: 18.38397789001465
Nearest Class Center Accuracy: 0.9949

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 68.80902862548828
Linear Weight Rank: 4031
Intra Cos: 0.9310899376869202
Inter Cos: 0.10637161135673523
Norm Quadratic Average: 122.8812255859375
Nearest Class Center Accuracy: 0.9946

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 27.94831085205078
Linear Weight Rank: 3670
Intra Cos: 0.9630906581878662
Inter Cos: 0.0604010708630085
Norm Quadratic Average: 64.19963073730469
Nearest Class Center Accuracy: 0.9954

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.4492595195770264
Linear Weight Rank: 10
Intra Cos: 0.9619066715240479
Inter Cos: 0.05880056694149971
Norm Quadratic Average: 31.0672664642334
Nearest Class Center Accuracy: 0.9956

Output Layer:
Intra Cos: 0.9873064160346985
Inter Cos: 0.3110692799091339
Norm Quadratic Average: 18.81134605407715
Nearest Class Center Accuracy: 0.9956

