Model save path: ./New_Models/bn_True_dataset_CIFAR10_epochs_200_lr_0.01_model_type_vgg19_rand_seed_979323_test_samples_None_train_samples_None_weight_decay_0.01.pth.tar
Training Set:
Layer 0: Conv2d(3, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.0198909193277359
Inter Cos: 0.10477277636528015
Norm Quadratic Average: 27.597179412841797
Nearest Class Center Accuracy: 0.26968

Layer 1: Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.012120941653847694
Inter Cos: 0.0606251135468483
Norm Quadratic Average: 6.317917346954346
Nearest Class Center Accuracy: 0.3407

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.017699606716632843
Inter Cos: 0.06251262128353119
Norm Quadratic Average: 4.816734790802002
Nearest Class Center Accuracy: 0.47084

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.012852108106017113
Inter Cos: 0.04869004711508751
Norm Quadratic Average: 3.5355443954467773
Nearest Class Center Accuracy: 0.55068

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.014090916141867638
Inter Cos: 0.043527472764253616
Norm Quadratic Average: 2.4451684951782227
Nearest Class Center Accuracy: 0.6329

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.01660045050084591
Inter Cos: 0.03477641940116882
Norm Quadratic Average: 1.8149588108062744
Nearest Class Center Accuracy: 0.69672

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.02053307369351387
Inter Cos: 0.03445453941822052
Norm Quadratic Average: 1.5037552118301392
Nearest Class Center Accuracy: 0.75996

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.026306042447686195
Inter Cos: 0.028760457411408424
Norm Quadratic Average: 1.2845385074615479
Nearest Class Center Accuracy: 0.8398

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.07510899752378464
Inter Cos: 0.052044615149497986
Norm Quadratic Average: 0.898444414138794
Nearest Class Center Accuracy: 0.94106

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.21157194674015045
Inter Cos: 0.07915159314870834
Norm Quadratic Average: 0.6037156581878662
Nearest Class Center Accuracy: 0.95808

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4722483158111572
Inter Cos: 0.10548748075962067
Norm Quadratic Average: 0.42761215567588806
Nearest Class Center Accuracy: 0.97854

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5780332684516907
Inter Cos: 0.047079797834157944
Norm Quadratic Average: 0.257948100566864
Nearest Class Center Accuracy: 0.9992

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8054284453392029
Inter Cos: 0.040667660534381866
Norm Quadratic Average: 0.18519045412540436
Nearest Class Center Accuracy: 0.99998

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9334549307823181
Inter Cos: 0.0991545170545578
Norm Quadratic Average: 0.19648025929927826
Nearest Class Center Accuracy: 0.99998

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9889963269233704
Inter Cos: 0.004790235310792923
Norm Quadratic Average: 0.44749724864959717
Nearest Class Center Accuracy: 0.99998

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9970185160636902
Inter Cos: -0.050148800015449524
Norm Quadratic Average: 1.0204861164093018
Nearest Class Center Accuracy: 0.99998

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.0928666591644287
Linear Weight Rank: 10
Intra Cos: 0.9990012049674988
Inter Cos: -0.02208852395415306
Norm Quadratic Average: 25.046066284179688
Nearest Class Center Accuracy: 0.99998

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.0976321697235107
Linear Weight Rank: 1346
Intra Cos: 0.9991589188575745
Inter Cos: 0.0060427384451031685
Norm Quadratic Average: 16.40129852294922
Nearest Class Center Accuracy: 0.99998

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.0958268642425537
Linear Weight Rank: 9
Intra Cos: 0.9992232918739319
Inter Cos: 0.06971312314271927
Norm Quadratic Average: 10.932510375976562
Nearest Class Center Accuracy: 0.99998

Output Layer:
Intra Cos: 0.9993672966957092
Inter Cos: 0.1603475958108902
Norm Quadratic Average: 7.712382793426514
Nearest Class Center Accuracy: 0.99998

Test Set:
Average Loss: 0.4424657826602459
Accuracy: 0.8949
NC1 Within Class Collapse: 2.0615813732147217
NC2 Equinorm: Features: 0.08110228925943375, Weights: 0.006762278266251087
NC2 Equiangle: Features: 0.09282734129163954, Weights: 0.03268292744954427
NC3 Self-Duality: 0.018694816157221794
NC4 NCC Mismatch: 0.00539999999999996

Layer 0: Conv2d(3, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.018526077270507812
Inter Cos: 0.1042705774307251
Norm Quadratic Average: 27.550132751464844
Nearest Class Center Accuracy: 0.284

Layer 1: Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.011168538592755795
Inter Cos: 0.061716821044683456
Norm Quadratic Average: 6.313758373260498
Nearest Class Center Accuracy: 0.3608

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.016484210267663002
Inter Cos: 0.06393565237522125
Norm Quadratic Average: 4.815937519073486
Nearest Class Center Accuracy: 0.4903

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.0120480265468359
Inter Cos: 0.049471043050289154
Norm Quadratic Average: 3.5359318256378174
Nearest Class Center Accuracy: 0.5736

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.013146992772817612
Inter Cos: 0.04436569660902023
Norm Quadratic Average: 2.446613311767578
Nearest Class Center Accuracy: 0.6448

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.015070118941366673
Inter Cos: 0.0354398675262928
Norm Quadratic Average: 1.8162810802459717
Nearest Class Center Accuracy: 0.6898

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.017654867842793465
Inter Cos: 0.03505571559071541
Norm Quadratic Average: 1.5025873184204102
Nearest Class Center Accuracy: 0.7328

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.021016934886574745
Inter Cos: 0.029778646305203438
Norm Quadratic Average: 1.278865933418274
Nearest Class Center Accuracy: 0.7838

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.05261858180165291
Inter Cos: 0.05565556883811951
Norm Quadratic Average: 0.8874309062957764
Nearest Class Center Accuracy: 0.8434

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.18263858556747437
Inter Cos: 0.09415105730295181
Norm Quadratic Average: 0.5846861600875854
Nearest Class Center Accuracy: 0.8391

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3807952105998993
Inter Cos: 0.19614054262638092
Norm Quadratic Average: 0.40945470333099365
Nearest Class Center Accuracy: 0.8336

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4220186769962311
Inter Cos: 0.1598517894744873
Norm Quadratic Average: 0.24460271000862122
Nearest Class Center Accuracy: 0.8773

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5686929821968079
Inter Cos: 0.15712684392929077
Norm Quadratic Average: 0.17503449320793152
Nearest Class Center Accuracy: 0.8841

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6387714147567749
Inter Cos: 0.0867089256644249
Norm Quadratic Average: 0.18612760305404663
Nearest Class Center Accuracy: 0.8938

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6698712110519409
Inter Cos: 0.09209291636943817
Norm Quadratic Average: 0.4218364357948303
Nearest Class Center Accuracy: 0.8945

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6774237155914307
Inter Cos: 0.09141331166028976
Norm Quadratic Average: 0.9584850072860718
Nearest Class Center Accuracy: 0.8946

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.0928666591644287
Linear Weight Rank: 10
Intra Cos: 0.6720055341720581
Inter Cos: 0.10835549235343933
Norm Quadratic Average: 23.466615676879883
Nearest Class Center Accuracy: 0.8945

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.0976321697235107
Linear Weight Rank: 1346
Intra Cos: 0.6795743703842163
Inter Cos: 0.15476669371128082
Norm Quadratic Average: 15.38216495513916
Nearest Class Center Accuracy: 0.8943

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.0958268642425537
Linear Weight Rank: 9
Intra Cos: 0.6885672807693481
Inter Cos: 0.20735900104045868
Norm Quadratic Average: 10.265722274780273
Nearest Class Center Accuracy: 0.8943

Output Layer:
Intra Cos: 0.701465368270874
Inter Cos: 0.26472777128219604
Norm Quadratic Average: 7.2565531730651855
Nearest Class Center Accuracy: 0.8941

