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_0.001.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: 0.054901983588933945
Inter Cos: 0.06699033826589584
Norm Quadratic Average: 7.071117401123047
Nearest Class Center Accuracy: 0.81855

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.09521506726741791
Inter Cos: 0.0866006389260292
Norm Quadratic Average: 4.669130325317383
Nearest Class Center Accuracy: 0.8789666666666667

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.09045226126909256
Inter Cos: 0.08081556856632233
Norm Quadratic Average: 4.013850212097168
Nearest Class Center Accuracy: 0.8889333333333334

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.16188718378543854
Inter Cos: 0.10954231023788452
Norm Quadratic Average: 3.1999337673187256
Nearest Class Center Accuracy: 0.9352833333333334

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.20021413266658783
Inter Cos: 0.12059814482927322
Norm Quadratic Average: 2.267922878265381
Nearest Class Center Accuracy: 0.9594333333333334

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.24355384707450867
Inter Cos: 0.12146691232919693
Norm Quadratic Average: 2.1515471935272217
Nearest Class Center Accuracy: 0.9742

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.28007277846336365
Inter Cos: 0.12543894350528717
Norm Quadratic Average: 2.0320332050323486
Nearest Class Center Accuracy: 0.98125

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.37230464816093445
Inter Cos: 0.12883220613002777
Norm Quadratic Average: 1.6210179328918457
Nearest Class Center Accuracy: 0.9952

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5733862519264221
Inter Cos: 0.15022313594818115
Norm Quadratic Average: 1.156255841255188
Nearest Class Center Accuracy: 0.9986333333333334

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7403189539909363
Inter Cos: 0.1347801685333252
Norm Quadratic Average: 1.0453871488571167
Nearest Class Center Accuracy: 0.9994

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8253543376922607
Inter Cos: 0.05180077999830246
Norm Quadratic Average: 0.8760793209075928
Nearest Class Center Accuracy: 0.9998666666666667

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8933537602424622
Inter Cos: 0.048997912555933
Norm Quadratic Average: 0.72367262840271
Nearest Class Center Accuracy: 0.9999833333333333

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.976996123790741
Inter Cos: -0.02333754114806652
Norm Quadratic Average: 0.6014701128005981
Nearest Class Center Accuracy: 1.0

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9956039190292358
Inter Cos: -0.028468405827879906
Norm Quadratic Average: 0.6583034992218018
Nearest Class Center Accuracy: 1.0

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9986962080001831
Inter Cos: -0.04038875177502632
Norm Quadratic Average: 1.0674633979797363
Nearest Class Center Accuracy: 1.0

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 3.088945150375366
Linear Weight Rank: 4028
Intra Cos: 0.9995437860488892
Inter Cos: -0.041721411049366
Norm Quadratic Average: 26.41073226928711
Nearest Class Center Accuracy: 1.0

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.4749574661254883
Linear Weight Rank: 3637
Intra Cos: 0.9995552897453308
Inter Cos: 0.013353344053030014
Norm Quadratic Average: 18.757863998413086
Nearest Class Center Accuracy: 1.0

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.2964656352996826
Linear Weight Rank: 9
Intra Cos: 0.9995114207267761
Inter Cos: 0.03446601331233978
Norm Quadratic Average: 13.67977237701416
Nearest Class Center Accuracy: 1.0

Output Layer:
Intra Cos: 0.9998319745063782
Inter Cos: 0.0975487232208252
Norm Quadratic Average: 10.776384353637695
Nearest Class Center Accuracy: 1.0

Test Set:
Average Loss: 0.02096150624498259
Accuracy: 0.9963
NC1 Within Class Collapse: 0.07212431728839874
NC2 Equinorm: Features: 0.011018001474440098, Weights: 0.009544393979012966
NC2 Equiangle: Features: 0.0863979975382487, Weights: 0.05792318979899089
NC3 Self-Duality: 0.01443109754472971
NC4 NCC Mismatch: 0.00019999999999997797

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: 0.06221870332956314
Inter Cos: 0.06871714442968369
Norm Quadratic Average: 7.044643402099609
Nearest Class Center Accuracy: 0.8304

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10416484624147415
Inter Cos: 0.08837936818599701
Norm Quadratic Average: 4.633936882019043
Nearest Class Center Accuracy: 0.8866

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.09872875362634659
Inter Cos: 0.08284486830234528
Norm Quadratic Average: 3.993830680847168
Nearest Class Center Accuracy: 0.8964

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.17381861805915833
Inter Cos: 0.11660394072532654
Norm Quadratic Average: 3.1836862564086914
Nearest Class Center Accuracy: 0.9421

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.21406644582748413
Inter Cos: 0.1321706771850586
Norm Quadratic Average: 2.259399175643921
Nearest Class Center Accuracy: 0.9613

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.25883495807647705
Inter Cos: 0.13003480434417725
Norm Quadratic Average: 2.1463840007781982
Nearest Class Center Accuracy: 0.9728

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2954334616661072
Inter Cos: 0.12181425839662552
Norm Quadratic Average: 2.027358055114746
Nearest Class Center Accuracy: 0.9788

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.38532185554504395
Inter Cos: 0.13986200094223022
Norm Quadratic Average: 1.6176353693008423
Nearest Class Center Accuracy: 0.9905

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5828603506088257
Inter Cos: 0.15870870649814606
Norm Quadratic Average: 1.153687834739685
Nearest Class Center Accuracy: 0.9935

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.746030330657959
Inter Cos: 0.13988520205020905
Norm Quadratic Average: 1.04378342628479
Nearest Class Center Accuracy: 0.9948

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8276196718215942
Inter Cos: 0.05780930817127228
Norm Quadratic Average: 0.8753854632377625
Nearest Class Center Accuracy: 0.9956

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8915413618087769
Inter Cos: 0.04848543182015419
Norm Quadratic Average: 0.7232067584991455
Nearest Class Center Accuracy: 0.9962

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9690566658973694
Inter Cos: -0.02527361735701561
Norm Quadratic Average: 0.6004471778869629
Nearest Class Center Accuracy: 0.9962

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9795407652854919
Inter Cos: -0.02993578277528286
Norm Quadratic Average: 0.6566526293754578
Nearest Class Center Accuracy: 0.9963

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.983214259147644
Inter Cos: -0.041637759655714035
Norm Quadratic Average: 1.0646308660507202
Nearest Class Center Accuracy: 0.9963

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 3.088945150375366
Linear Weight Rank: 4028
Intra Cos: 0.984575092792511
Inter Cos: -0.03819187358021736
Norm Quadratic Average: 26.339191436767578
Nearest Class Center Accuracy: 0.9962

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.4749574661254883
Linear Weight Rank: 3637
Intra Cos: 0.9853951930999756
Inter Cos: 0.01520299818366766
Norm Quadratic Average: 18.707958221435547
Nearest Class Center Accuracy: 0.9962

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.2964656352996826
Linear Weight Rank: 9
Intra Cos: 0.9858766794204712
Inter Cos: 0.04369359463453293
Norm Quadratic Average: 13.645068168640137
Nearest Class Center Accuracy: 0.9963

Output Layer:
Intra Cos: 0.9875330924987793
Inter Cos: 0.10575108230113983
Norm Quadratic Average: 10.749755859375
Nearest Class Center Accuracy: 0.9964

