Model save path: ./New_Models/bn_False_dataset_MNIST_epochs_200_lr_0.01_model_type_vgg19_rand_seed_314159_test_samples_None_train_samples_None_weight_decay_0.0003.pth.tar
Training Set:
Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.09116753935813904
Inter Cos: 0.10967151075601578
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: 0.08943930268287659
Inter Cos: 0.10859601199626923
Norm Quadratic Average: 58.20436477661133
Nearest Class Center Accuracy: 0.81415

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.12614277005195618
Inter Cos: 0.1392129510641098
Norm Quadratic Average: 65.88107299804688
Nearest Class Center Accuracy: 0.83505

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1341371238231659
Inter Cos: 0.15057356655597687
Norm Quadratic Average: 100.42401885986328
Nearest Class Center Accuracy: 0.8461666666666666

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.20124401152133942
Inter Cos: 0.17789624631404877
Norm Quadratic Average: 79.90652465820312
Nearest Class Center Accuracy: 0.89775

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2343195080757141
Inter Cos: 0.1850673407316208
Norm Quadratic Average: 78.84735107421875
Nearest Class Center Accuracy: 0.92425

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.26305055618286133
Inter Cos: 0.1793501228094101
Norm Quadratic Average: 69.08513641357422
Nearest Class Center Accuracy: 0.94095

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.28381866216659546
Inter Cos: 0.1736590564250946
Norm Quadratic Average: 54.555442810058594
Nearest Class Center Accuracy: 0.95

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.30430418252944946
Inter Cos: 0.1595742255449295
Norm Quadratic Average: 23.709444046020508
Nearest Class Center Accuracy: 0.9757166666666667

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.45404812693595886
Inter Cos: 0.1843462437391281
Norm Quadratic Average: 16.27317237854004
Nearest Class Center Accuracy: 0.9884166666666667

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5557245016098022
Inter Cos: 0.2325764298439026
Norm Quadratic Average: 14.082908630371094
Nearest Class Center Accuracy: 0.9928666666666667

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5946241021156311
Inter Cos: 0.23166383802890778
Norm Quadratic Average: 13.090655326843262
Nearest Class Center Accuracy: 0.9952333333333333

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6362305879592896
Inter Cos: 0.19570712745189667
Norm Quadratic Average: 7.986531734466553
Nearest Class Center Accuracy: 0.9953666666666666

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8512701392173767
Inter Cos: 0.22929085791110992
Norm Quadratic Average: 6.6723127365112305
Nearest Class Center Accuracy: 0.9969833333333333

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9075531363487244
Inter Cos: 0.2429986298084259
Norm Quadratic Average: 6.25093936920166
Nearest Class Center Accuracy: 0.9974333333333333

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9263268113136292
Inter Cos: 0.25879862904548645
Norm Quadratic Average: 5.8469038009643555
Nearest Class Center Accuracy: 0.9977666666666667

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 31.700706481933594
Linear Weight Rank: 4031
Intra Cos: 0.9389140605926514
Inter Cos: 0.2691519558429718
Norm Quadratic Average: 33.65799331665039
Nearest Class Center Accuracy: 0.9977666666666667

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 13.033122062683105
Linear Weight Rank: 3669
Intra Cos: 0.9460847973823547
Inter Cos: 0.2849656045436859
Norm Quadratic Average: 29.212705612182617
Nearest Class Center Accuracy: 0.9978

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.6109471321105957
Linear Weight Rank: 10
Intra Cos: 0.9490105509757996
Inter Cos: 0.24591219425201416
Norm Quadratic Average: 25.671470642089844
Nearest Class Center Accuracy: 0.99815

Output Layer:
Intra Cos: 0.97953861951828
Inter Cos: 0.25540944933891296
Norm Quadratic Average: 24.79447364807129
Nearest Class Center Accuracy: 0.99985

Test Set:
Average Loss: 0.024403769741432917
Accuracy: 0.9939
NC1 Within Class Collapse: 0.541864812374115
NC2 Equinorm: Features: 0.10407370328903198, Weights: 0.06343776732683182
NC2 Equiangle: Features: 0.23663645850287543, Weights: 0.1489983876546224
NC3 Self-Duality: 0.206088125705719
NC4 NCC Mismatch: 0.0046000000000000485

Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10219582915306091
Inter Cos: 0.12048851698637009
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: 0.10014740377664566
Inter Cos: 0.1172192245721817
Norm Quadratic Average: 58.27021408081055
Nearest Class Center Accuracy: 0.8269

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.13966436684131622
Inter Cos: 0.1526772528886795
Norm Quadratic Average: 65.75731658935547
Nearest Class Center Accuracy: 0.8492

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.14759871363639832
Inter Cos: 0.1645924597978592
Norm Quadratic Average: 100.33351135253906
Nearest Class Center Accuracy: 0.8588

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.21849843859672546
Inter Cos: 0.1929539144039154
Norm Quadratic Average: 79.77452850341797
Nearest Class Center Accuracy: 0.908

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.252124547958374
Inter Cos: 0.19188308715820312
Norm Quadratic Average: 78.7741928100586
Nearest Class Center Accuracy: 0.9351

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.28119778633117676
Inter Cos: 0.18634088337421417
Norm Quadratic Average: 69.06900787353516
Nearest Class Center Accuracy: 0.9486

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3008947968482971
Inter Cos: 0.19150613248348236
Norm Quadratic Average: 54.58539581298828
Nearest Class Center Accuracy: 0.9558

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3155408799648285
Inter Cos: 0.17591603100299835
Norm Quadratic Average: 23.72607421875
Nearest Class Center Accuracy: 0.9764

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4691349267959595
Inter Cos: 0.20197264850139618
Norm Quadratic Average: 16.31488800048828
Nearest Class Center Accuracy: 0.9855

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5678223967552185
Inter Cos: 0.23239991068840027
Norm Quadratic Average: 14.141895294189453
Nearest Class Center Accuracy: 0.9883

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6039742827415466
Inter Cos: 0.24162480235099792
Norm Quadratic Average: 13.157636642456055
Nearest Class Center Accuracy: 0.9907

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6376305818557739
Inter Cos: 0.21073240041732788
Norm Quadratic Average: 8.032120704650879
Nearest Class Center Accuracy: 0.9897

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8507775664329529
Inter Cos: 0.24680744111537933
Norm Quadratic Average: 6.716579437255859
Nearest Class Center Accuracy: 0.9898

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9063304662704468
Inter Cos: 0.25674667954444885
Norm Quadratic Average: 6.29084587097168
Nearest Class Center Accuracy: 0.9902

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9248557090759277
Inter Cos: 0.2562873065471649
Norm Quadratic Average: 5.882862567901611
Nearest Class Center Accuracy: 0.9904

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 31.700706481933594
Linear Weight Rank: 4031
Intra Cos: 0.936547040939331
Inter Cos: 0.26434165239334106
Norm Quadratic Average: 33.86486053466797
Nearest Class Center Accuracy: 0.9908

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 13.033122062683105
Linear Weight Rank: 3669
Intra Cos: 0.9436505436897278
Inter Cos: 0.27927732467651367
Norm Quadratic Average: 29.39005470275879
Nearest Class Center Accuracy: 0.9912

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.6109471321105957
Linear Weight Rank: 10
Intra Cos: 0.9453377723693848
Inter Cos: 0.24105851352214813
Norm Quadratic Average: 25.828596115112305
Nearest Class Center Accuracy: 0.9915

Output Layer:
Intra Cos: 0.9692683815956116
Inter Cos: 0.2689566910266876
Norm Quadratic Average: 24.94419288635254
Nearest Class Center Accuracy: 0.9938

