Model save path: ./New_Models/bn_False_dataset_MNIST_epochs_200_lr_0.01_model_type_vgg19_rand_seed_265358_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: 0.09116753190755844
Inter Cos: 0.10967151820659637
Norm Quadratic Average: 23.567678451538086
Nearest Class Center Accuracy: 0.8079833333333334

Layer 1: Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10104059427976608
Inter Cos: 0.1259632706642151
Norm Quadratic Average: 61.84416580200195
Nearest Class Center Accuracy: 0.80745

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1259463131427765
Inter Cos: 0.16390398144721985
Norm Quadratic Average: 85.05027770996094
Nearest Class Center Accuracy: 0.8075333333333333

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.13316184282302856
Inter Cos: 0.17888127267360687
Norm Quadratic Average: 148.61497497558594
Nearest Class Center Accuracy: 0.8060333333333334

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.18981099128723145
Inter Cos: 0.1995706856250763
Norm Quadratic Average: 129.4412384033203
Nearest Class Center Accuracy: 0.83715

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2185819298028946
Inter Cos: 0.21339064836502075
Norm Quadratic Average: 134.60462951660156
Nearest Class Center Accuracy: 0.85875

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.23477408289909363
Inter Cos: 0.2254026234149933
Norm Quadratic Average: 142.20191955566406
Nearest Class Center Accuracy: 0.8808166666666667

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.25824880599975586
Inter Cos: 0.22280503809452057
Norm Quadratic Average: 122.09051513671875
Nearest Class Center Accuracy: 0.9083

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.24839404225349426
Inter Cos: 0.19986872375011444
Norm Quadratic Average: 47.67902374267578
Nearest Class Center Accuracy: 0.94475

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3822503685951233
Inter Cos: 0.22402402758598328
Norm Quadratic Average: 23.199853897094727
Nearest Class Center Accuracy: 0.9721

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.49236178398132324
Inter Cos: 0.27664610743522644
Norm Quadratic Average: 16.890338897705078
Nearest Class Center Accuracy: 0.9801833333333333

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5782855749130249
Inter Cos: 0.2956719398498535
Norm Quadratic Average: 16.105209350585938
Nearest Class Center Accuracy: 0.9863166666666666

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5711519718170166
Inter Cos: 0.3636515438556671
Norm Quadratic Average: 9.831663131713867
Nearest Class Center Accuracy: 0.9856666666666667

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7970718741416931
Inter Cos: 0.4571237862110138
Norm Quadratic Average: 8.685298919677734
Nearest Class Center Accuracy: 0.9905333333333334

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8599602580070496
Inter Cos: 0.458208292722702
Norm Quadratic Average: 9.557759284973145
Nearest Class Center Accuracy: 0.9942

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.884941577911377
Inter Cos: 0.44414380192756653
Norm Quadratic Average: 9.99396800994873
Nearest Class Center Accuracy: 0.9957166666666667

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.8865604400634766
Linear Weight Rank: 4029
Intra Cos: 0.8895426392555237
Inter Cos: 0.3070546090602875
Norm Quadratic Average: 47.396427154541016
Nearest Class Center Accuracy: 0.9966666666666667

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.264021635055542
Linear Weight Rank: 3641
Intra Cos: 0.9191181063652039
Inter Cos: 0.28028249740600586
Norm Quadratic Average: 37.250247955322266
Nearest Class Center Accuracy: 0.9982833333333333

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.0311501026153564
Linear Weight Rank: 9
Intra Cos: 0.9274976253509521
Inter Cos: 0.30060288310050964
Norm Quadratic Average: 28.297897338867188
Nearest Class Center Accuracy: 0.9988

Output Layer:
Intra Cos: 0.9673100709915161
Inter Cos: 0.3980509340763092
Norm Quadratic Average: 24.070404052734375
Nearest Class Center Accuracy: 0.9997833333333334

Test Set:
Average Loss: 0.02846680366671644
Accuracy: 0.9917
NC1 Within Class Collapse: 1.638479232788086
NC2 Equinorm: Features: 0.10091416537761688, Weights: 0.037085533142089844
NC2 Equiangle: Features: 0.2634635077582465, Weights: 0.20772370232476128
NC3 Self-Duality: 0.0897451639175415
NC4 NCC Mismatch: 0.0044999999999999485

Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10219582915306091
Inter Cos: 0.12048853933811188
Norm Quadratic Average: 23.595195770263672
Nearest Class Center Accuracy: 0.8229

Layer 1: Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.11276867985725403
Inter Cos: 0.13859881460666656
Norm Quadratic Average: 62.07212829589844
Nearest Class Center Accuracy: 0.8205

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.14117875695228577
Inter Cos: 0.17955800890922546
Norm Quadratic Average: 85.1958999633789
Nearest Class Center Accuracy: 0.8234

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.14890050888061523
Inter Cos: 0.1959296315908432
Norm Quadratic Average: 148.94973754882812
Nearest Class Center Accuracy: 0.8197

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.20712676644325256
Inter Cos: 0.21818353235721588
Norm Quadratic Average: 129.47137451171875
Nearest Class Center Accuracy: 0.8518

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.23931972682476044
Inter Cos: 0.2340887188911438
Norm Quadratic Average: 134.5886993408203
Nearest Class Center Accuracy: 0.8751

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.255149245262146
Inter Cos: 0.24705594778060913
Norm Quadratic Average: 142.2614288330078
Nearest Class Center Accuracy: 0.8943

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2773498296737671
Inter Cos: 0.24326954782009125
Norm Quadratic Average: 122.42355346679688
Nearest Class Center Accuracy: 0.9182

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2646283805370331
Inter Cos: 0.21864372491836548
Norm Quadratic Average: 47.883548736572266
Nearest Class Center Accuracy: 0.9508

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.39881670475006104
Inter Cos: 0.24368464946746826
Norm Quadratic Average: 23.338050842285156
Nearest Class Center Accuracy: 0.9735

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5060827136039734
Inter Cos: 0.27232927083969116
Norm Quadratic Average: 17.018274307250977
Nearest Class Center Accuracy: 0.9779

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5846818089485168
Inter Cos: 0.2959025502204895
Norm Quadratic Average: 16.237089157104492
Nearest Class Center Accuracy: 0.9811

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5725204348564148
Inter Cos: 0.38055935502052307
Norm Quadratic Average: 9.911718368530273
Nearest Class Center Accuracy: 0.9792

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7998139262199402
Inter Cos: 0.46766436100006104
Norm Quadratic Average: 8.776326179504395
Nearest Class Center Accuracy: 0.9829

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.860742449760437
Inter Cos: 0.4646262228488922
Norm Quadratic Average: 9.668624877929688
Nearest Class Center Accuracy: 0.9856

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8873529434204102
Inter Cos: 0.4485090672969818
Norm Quadratic Average: 10.112412452697754
Nearest Class Center Accuracy: 0.9872

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.8865604400634766
Linear Weight Rank: 4029
Intra Cos: 0.8901259899139404
Inter Cos: 0.31130045652389526
Norm Quadratic Average: 47.929847717285156
Nearest Class Center Accuracy: 0.9883

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.264021635055542
Linear Weight Rank: 3641
Intra Cos: 0.9190904498100281
Inter Cos: 0.29652923345565796
Norm Quadratic Average: 37.686439514160156
Nearest Class Center Accuracy: 0.9899

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.0311501026153564
Linear Weight Rank: 9
Intra Cos: 0.9254239201545715
Inter Cos: 0.316641241312027
Norm Quadratic Average: 28.640378952026367
Nearest Class Center Accuracy: 0.9904

Output Layer:
Intra Cos: 0.9580051302909851
Inter Cos: 0.41330721974372864
Norm Quadratic Average: 24.3700008392334
Nearest Class Center Accuracy: 0.9913

