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_0.01.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: 0.11909449845552444
Inter Cos: 0.14346294105052948
Norm Quadratic Average: 67.00520324707031
Nearest Class Center Accuracy: 0.7991166666666667

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1415150910615921
Inter Cos: 0.1845511496067047
Norm Quadratic Average: 132.1894073486328
Nearest Class Center Accuracy: 0.7739833333333334

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.14697912335395813
Inter Cos: 0.19526244699954987
Norm Quadratic Average: 262.41619873046875
Nearest Class Center Accuracy: 0.7679666666666667

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.16606692969799042
Inter Cos: 0.19641712307929993
Norm Quadratic Average: 193.12738037109375
Nearest Class Center Accuracy: 0.7739333333333334

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1755370795726776
Inter Cos: 0.20895205438137054
Norm Quadratic Average: 166.31910705566406
Nearest Class Center Accuracy: 0.7711666666666667

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.18680286407470703
Inter Cos: 0.22942572832107544
Norm Quadratic Average: 165.29550170898438
Nearest Class Center Accuracy: 0.8072

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.22009815275669098
Inter Cos: 0.2644609212875366
Norm Quadratic Average: 168.30711364746094
Nearest Class Center Accuracy: 0.8672166666666666

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.17499227821826935
Inter Cos: 0.2552354335784912
Norm Quadratic Average: 85.68692016601562
Nearest Class Center Accuracy: 0.8526666666666667

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1795881986618042
Inter Cos: 0.3544701337814331
Norm Quadratic Average: 51.984195709228516
Nearest Class Center Accuracy: 0.82425

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2650105655193329
Inter Cos: 0.3461952805519104
Norm Quadratic Average: 47.38438034057617
Nearest Class Center Accuracy: 0.8637666666666667

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4038273096084595
Inter Cos: 0.34496206045150757
Norm Quadratic Average: 46.13594055175781
Nearest Class Center Accuracy: 0.9184166666666667

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.45199593901634216
Inter Cos: 0.31701400876045227
Norm Quadratic Average: 25.956668853759766
Nearest Class Center Accuracy: 0.9059166666666667

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5769195556640625
Inter Cos: 0.41769593954086304
Norm Quadratic Average: 18.81546401977539
Nearest Class Center Accuracy: 0.9231166666666667

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6396093368530273
Inter Cos: 0.4304470121860504
Norm Quadratic Average: 19.15020751953125
Nearest Class Center Accuracy: 0.9559333333333333

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6870617866516113
Inter Cos: 0.42679789662361145
Norm Quadratic Average: 20.33115577697754
Nearest Class Center Accuracy: 0.9732166666666666

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 1.5169485807418823
Linear Weight Rank: 11
Intra Cos: 0.7701756954193115
Inter Cos: 0.3830508589744568
Norm Quadratic Average: 82.95121765136719
Nearest Class Center Accuracy: 0.9834666666666667

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 1.5237057209014893
Linear Weight Rank: 2754
Intra Cos: 0.8486547470092773
Inter Cos: 0.37754493951797485
Norm Quadratic Average: 50.37965393066406
Nearest Class Center Accuracy: 0.9905333333333334

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 1.5137181282043457
Linear Weight Rank: 9
Intra Cos: 0.875889241695404
Inter Cos: 0.3510991632938385
Norm Quadratic Average: 28.70488929748535
Nearest Class Center Accuracy: 0.9922333333333333

Output Layer:
Intra Cos: 0.9203649163246155
Inter Cos: 0.47467660903930664
Norm Quadratic Average: 18.095441818237305
Nearest Class Center Accuracy: 0.9935833333333334

Test Set:
Average Loss: 0.044416343054547905
Accuracy: 0.987
NC1 Within Class Collapse: 1.7033631801605225
NC2 Equinorm: Features: 0.1086430698633194, Weights: 0.0349557064473629
NC2 Equiangle: Features: 0.2896662394205729, Weights: 0.21457057529025608
NC3 Self-Duality: 0.08277924358844757
NC4 NCC Mismatch: 0.007000000000000006

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: 0.13084155321121216
Inter Cos: 0.1570604145526886
Norm Quadratic Average: 67.42910766601562
Nearest Class Center Accuracy: 0.8152

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1560966521501541
Inter Cos: 0.20207002758979797
Norm Quadratic Average: 132.8891143798828
Nearest Class Center Accuracy: 0.7948

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1615677922964096
Inter Cos: 0.21433813869953156
Norm Quadratic Average: 263.8348083496094
Nearest Class Center Accuracy: 0.7891

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.17721323668956757
Inter Cos: 0.21776384115219116
Norm Quadratic Average: 193.6672821044922
Nearest Class Center Accuracy: 0.796

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.18747782707214355
Inter Cos: 0.23244896531105042
Norm Quadratic Average: 166.72573852539062
Nearest Class Center Accuracy: 0.7939

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.19913484156131744
Inter Cos: 0.2548605799674988
Norm Quadratic Average: 165.77452087402344
Nearest Class Center Accuracy: 0.8317

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.23281973600387573
Inter Cos: 0.29115766286849976
Norm Quadratic Average: 169.22840881347656
Nearest Class Center Accuracy: 0.8806

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.18066681921482086
Inter Cos: 0.2794528603553772
Norm Quadratic Average: 86.27803039550781
Nearest Class Center Accuracy: 0.8635

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.17701315879821777
Inter Cos: 0.3845036029815674
Norm Quadratic Average: 52.34645462036133
Nearest Class Center Accuracy: 0.8363

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2628028094768524
Inter Cos: 0.37463319301605225
Norm Quadratic Average: 47.73202896118164
Nearest Class Center Accuracy: 0.867

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.40947479009628296
Inter Cos: 0.37070232629776
Norm Quadratic Average: 46.54359817504883
Nearest Class Center Accuracy: 0.9178

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4652325510978699
Inter Cos: 0.34201598167419434
Norm Quadratic Average: 26.16862678527832
Nearest Class Center Accuracy: 0.9098

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5837304592132568
Inter Cos: 0.4436156451702118
Norm Quadratic Average: 19.0321102142334
Nearest Class Center Accuracy: 0.9245

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.640418529510498
Inter Cos: 0.4531812071800232
Norm Quadratic Average: 19.420852661132812
Nearest Class Center Accuracy: 0.9529

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6839358806610107
Inter Cos: 0.44636568427085876
Norm Quadratic Average: 20.653961181640625
Nearest Class Center Accuracy: 0.9688

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 1.5169485807418823
Linear Weight Rank: 11
Intra Cos: 0.765684962272644
Inter Cos: 0.39780911803245544
Norm Quadratic Average: 84.34860229492188
Nearest Class Center Accuracy: 0.9764

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 1.5237057209014893
Linear Weight Rank: 2754
Intra Cos: 0.8390364050865173
Inter Cos: 0.3884817361831665
Norm Quadratic Average: 51.20878601074219
Nearest Class Center Accuracy: 0.982

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 1.5137181282043457
Linear Weight Rank: 9
Intra Cos: 0.8652990460395813
Inter Cos: 0.3471376597881317
Norm Quadratic Average: 29.178516387939453
Nearest Class Center Accuracy: 0.9837

Output Layer:
Intra Cos: 0.9094129800796509
Inter Cos: 0.4711664021015167
Norm Quadratic Average: 18.398794174194336
Nearest Class Center Accuracy: 0.9849

