Model save path: ./New_Models/bn_False_dataset_MNIST_epochs_200_lr_0.01_model_type_vgg19_rand_seed_979323_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: 0.10967149585485458
Norm Quadratic Average: 23.567672729492188
Nearest Class Center Accuracy: 0.8079833333333334

Layer 1: Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10011455416679382
Inter Cos: 0.12257805466651917
Norm Quadratic Average: 60.64996337890625
Nearest Class Center Accuracy: 0.80755

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.125833660364151
Inter Cos: 0.1553535759449005
Norm Quadratic Average: 78.853271484375
Nearest Class Center Accuracy: 0.8118166666666666

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.13204632699489594
Inter Cos: 0.16450051963329315
Norm Quadratic Average: 121.38123321533203
Nearest Class Center Accuracy: 0.8206166666666667

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1904558539390564
Inter Cos: 0.18004924058914185
Norm Quadratic Average: 88.87674713134766
Nearest Class Center Accuracy: 0.8745

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.22287137806415558
Inter Cos: 0.18644621968269348
Norm Quadratic Average: 76.72227478027344
Nearest Class Center Accuracy: 0.9077166666666666

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.25125038623809814
Inter Cos: 0.20482538640499115
Norm Quadratic Average: 66.27970123291016
Nearest Class Center Accuracy: 0.9283

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.28748857975006104
Inter Cos: 0.18934577703475952
Norm Quadratic Average: 48.92679977416992
Nearest Class Center Accuracy: 0.9482166666666667

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3110671639442444
Inter Cos: 0.17948585748672485
Norm Quadratic Average: 18.518470764160156
Nearest Class Center Accuracy: 0.9637833333333333

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5122621059417725
Inter Cos: 0.24985048174858093
Norm Quadratic Average: 10.135443687438965
Nearest Class Center Accuracy: 0.9743166666666667

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5893477201461792
Inter Cos: 0.30779603123664856
Norm Quadratic Average: 10.549722671508789
Nearest Class Center Accuracy: 0.9808166666666667

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6445596218109131
Inter Cos: 0.3685465157032013
Norm Quadratic Average: 11.994627952575684
Nearest Class Center Accuracy: 0.98815

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6656932830810547
Inter Cos: 0.28562766313552856
Norm Quadratic Average: 7.588229656219482
Nearest Class Center Accuracy: 0.98335

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8355293273925781
Inter Cos: 0.30072101950645447
Norm Quadratic Average: 6.543471813201904
Nearest Class Center Accuracy: 0.9869666666666667

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8774082064628601
Inter Cos: 0.30566704273223877
Norm Quadratic Average: 7.671030521392822
Nearest Class Center Accuracy: 0.99175

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8938245177268982
Inter Cos: 0.3197686970233917
Norm Quadratic Average: 8.61751937866211
Nearest Class Center Accuracy: 0.9942333333333333

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.8872485160827637
Linear Weight Rank: 4029
Intra Cos: 0.8914497494697571
Inter Cos: 0.2712642550468445
Norm Quadratic Average: 43.894439697265625
Nearest Class Center Accuracy: 0.9961

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.2704858779907227
Linear Weight Rank: 3641
Intra Cos: 0.920628011226654
Inter Cos: 0.28422585129737854
Norm Quadratic Average: 36.627342224121094
Nearest Class Center Accuracy: 0.9983

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.037050724029541
Linear Weight Rank: 9
Intra Cos: 0.937278687953949
Inter Cos: 0.2673005759716034
Norm Quadratic Average: 28.11345100402832
Nearest Class Center Accuracy: 0.9994

Output Layer:
Intra Cos: 0.9701365828514099
Inter Cos: 0.3353402316570282
Norm Quadratic Average: 24.031530380249023
Nearest Class Center Accuracy: 0.9998833333333333

Test Set:
Average Loss: 0.028431857495708392
Accuracy: 0.9931
NC1 Within Class Collapse: 0.724366307258606
NC2 Equinorm: Features: 0.08073806762695312, Weights: 0.03642981871962547
NC2 Equiangle: Features: 0.30468499925401477, Weights: 0.2145660400390625
NC3 Self-Duality: 0.10788039863109589
NC4 NCC Mismatch: 0.0038000000000000256

Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10219583660364151
Inter Cos: 0.12048852443695068
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: 0.11168363690376282
Inter Cos: 0.13486172258853912
Norm Quadratic Average: 60.854007720947266
Nearest Class Center Accuracy: 0.8206

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.14043737947940826
Inter Cos: 0.1702277511358261
Norm Quadratic Average: 78.9436264038086
Nearest Class Center Accuracy: 0.8265

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.14652954041957855
Inter Cos: 0.18017540872097015
Norm Quadratic Average: 121.6170425415039
Nearest Class Center Accuracy: 0.8333

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.20644252002239227
Inter Cos: 0.1964661180973053
Norm Quadratic Average: 88.97187042236328
Nearest Class Center Accuracy: 0.8857

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2393878549337387
Inter Cos: 0.18771560490131378
Norm Quadratic Average: 76.82713317871094
Nearest Class Center Accuracy: 0.918

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2656845450401306
Inter Cos: 0.19743606448173523
Norm Quadratic Average: 66.39557647705078
Nearest Class Center Accuracy: 0.9375

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3008202612400055
Inter Cos: 0.18113595247268677
Norm Quadratic Average: 49.084598541259766
Nearest Class Center Accuracy: 0.9562

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3271518647670746
Inter Cos: 0.18680456280708313
Norm Quadratic Average: 18.586153030395508
Nearest Class Center Accuracy: 0.9681

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.537241518497467
Inter Cos: 0.2743062674999237
Norm Quadratic Average: 10.185079574584961
Nearest Class Center Accuracy: 0.971

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6126025915145874
Inter Cos: 0.3256138563156128
Norm Quadratic Average: 10.618800163269043
Nearest Class Center Accuracy: 0.9744

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6630342602729797
Inter Cos: 0.38479724526405334
Norm Quadratic Average: 12.088150978088379
Nearest Class Center Accuracy: 0.9817

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.667905330657959
Inter Cos: 0.29978135228157043
Norm Quadratic Average: 7.651206970214844
Nearest Class Center Accuracy: 0.9765

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8432610630989075
Inter Cos: 0.31262537837028503
Norm Quadratic Average: 6.615415096282959
Nearest Class Center Accuracy: 0.979

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8830286860466003
Inter Cos: 0.3088354468345642
Norm Quadratic Average: 7.75648832321167
Nearest Class Center Accuracy: 0.9832

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8973698019981384
Inter Cos: 0.3219859004020691
Norm Quadratic Average: 8.707270622253418
Nearest Class Center Accuracy: 0.986

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.8872485160827637
Linear Weight Rank: 4029
Intra Cos: 0.8941428065299988
Inter Cos: 0.2694978713989258
Norm Quadratic Average: 44.310218811035156
Nearest Class Center Accuracy: 0.9883

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.2704858779907227
Linear Weight Rank: 3641
Intra Cos: 0.9204445481300354
Inter Cos: 0.28616076707839966
Norm Quadratic Average: 36.97064208984375
Nearest Class Center Accuracy: 0.9901

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.037050724029541
Linear Weight Rank: 9
Intra Cos: 0.9343550801277161
Inter Cos: 0.2701905369758606
Norm Quadratic Average: 28.378890991210938
Nearest Class Center Accuracy: 0.9914

Output Layer:
Intra Cos: 0.9615743160247803
Inter Cos: 0.3526292145252228
Norm Quadratic Average: 24.258150100708008
Nearest Class Center Accuracy: 0.9926

