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_0.007.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: 0.11858338862657547
Inter Cos: 0.14225754141807556
Norm Quadratic Average: 69.26390075683594
Nearest Class Center Accuracy: 0.7991

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1407388150691986
Inter Cos: 0.18213723599910736
Norm Quadratic Average: 136.8151397705078
Nearest Class Center Accuracy: 0.7753

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.14413879811763763
Inter Cos: 0.18832093477249146
Norm Quadratic Average: 251.27713012695312
Nearest Class Center Accuracy: 0.7761666666666667

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1731126308441162
Inter Cos: 0.19324557483196259
Norm Quadratic Average: 151.60528564453125
Nearest Class Center Accuracy: 0.8198166666666666

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2106320708990097
Inter Cos: 0.21546712517738342
Norm Quadratic Average: 102.71833038330078
Nearest Class Center Accuracy: 0.8478

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.24255867302417755
Inter Cos: 0.24170054495334625
Norm Quadratic Average: 94.47447967529297
Nearest Class Center Accuracy: 0.8844333333333333

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2733458876609802
Inter Cos: 0.28014877438545227
Norm Quadratic Average: 97.14012145996094
Nearest Class Center Accuracy: 0.91885

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2531818449497223
Inter Cos: 0.2902185916900635
Norm Quadratic Average: 49.09890365600586
Nearest Class Center Accuracy: 0.9116333333333333

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.22529298067092896
Inter Cos: 0.3067009747028351
Norm Quadratic Average: 32.90847396850586
Nearest Class Center Accuracy: 0.8990333333333334

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.28519728779792786
Inter Cos: 0.3165520131587982
Norm Quadratic Average: 31.172529220581055
Nearest Class Center Accuracy: 0.9151833333333333

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4039021134376526
Inter Cos: 0.3477930426597595
Norm Quadratic Average: 32.02104568481445
Nearest Class Center Accuracy: 0.9414166666666667

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5055702328681946
Inter Cos: 0.3553538918495178
Norm Quadratic Average: 20.04132843017578
Nearest Class Center Accuracy: 0.93135

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.699219822883606
Inter Cos: 0.3960711658000946
Norm Quadratic Average: 15.64855670928955
Nearest Class Center Accuracy: 0.96685

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7511883974075317
Inter Cos: 0.41493093967437744
Norm Quadratic Average: 17.161447525024414
Nearest Class Center Accuracy: 0.97955

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7776841521263123
Inter Cos: 0.41972294449806213
Norm Quadratic Average: 19.360437393188477
Nearest Class Center Accuracy: 0.9846166666666667

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 1.5682508945465088
Linear Weight Rank: 18
Intra Cos: 0.8455725908279419
Inter Cos: 0.3486061096191406
Norm Quadratic Average: 80.76878356933594
Nearest Class Center Accuracy: 0.98855

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 1.5756114721298218
Linear Weight Rank: 2713
Intra Cos: 0.8922471404075623
Inter Cos: 0.34606704115867615
Norm Quadratic Average: 50.3970947265625
Nearest Class Center Accuracy: 0.9927666666666667

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 1.5645655393600464
Linear Weight Rank: 9
Intra Cos: 0.9076297879219055
Inter Cos: 0.32559022307395935
Norm Quadratic Average: 28.95611572265625
Nearest Class Center Accuracy: 0.994

Output Layer:
Intra Cos: 0.9483907222747803
Inter Cos: 0.3915771543979645
Norm Quadratic Average: 18.385141372680664
Nearest Class Center Accuracy: 0.9959666666666667

Test Set:
Average Loss: 0.03775216683894396
Accuracy: 0.9886
NC1 Within Class Collapse: 1.5607037544250488
NC2 Equinorm: Features: 0.08976756781339645, Weights: 0.05779575929045677
NC2 Equiangle: Features: 0.28343641493055555, Weights: 0.19687491522894965
NC3 Self-Duality: 0.0995677188038826
NC4 NCC Mismatch: 0.010199999999999987

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: 0.13025467097759247
Inter Cos: 0.1557362675666809
Norm Quadratic Average: 69.70189666748047
Nearest Class Center Accuracy: 0.8157

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.15508364140987396
Inter Cos: 0.19944998621940613
Norm Quadratic Average: 137.5299530029297
Nearest Class Center Accuracy: 0.7957

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.15840815007686615
Inter Cos: 0.2066701054573059
Norm Quadratic Average: 252.6093292236328
Nearest Class Center Accuracy: 0.7959

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.18452376127243042
Inter Cos: 0.21251411736011505
Norm Quadratic Average: 152.05760192871094
Nearest Class Center Accuracy: 0.8346

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.22499863803386688
Inter Cos: 0.23660078644752502
Norm Quadratic Average: 102.99885559082031
Nearest Class Center Accuracy: 0.8633

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.259458988904953
Inter Cos: 0.2634066939353943
Norm Quadratic Average: 94.72452545166016
Nearest Class Center Accuracy: 0.8991

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.28949031233787537
Inter Cos: 0.27544429898262024
Norm Quadratic Average: 97.69727325439453
Nearest Class Center Accuracy: 0.9295

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.27050337195396423
Inter Cos: 0.30962371826171875
Norm Quadratic Average: 49.40633010864258
Nearest Class Center Accuracy: 0.9183

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2319817841053009
Inter Cos: 0.3310062289237976
Norm Quadratic Average: 33.10557556152344
Nearest Class Center Accuracy: 0.9064

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2835601270198822
Inter Cos: 0.34659865498542786
Norm Quadratic Average: 31.40943717956543
Nearest Class Center Accuracy: 0.9186

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.39988988637924194
Inter Cos: 0.3804088234901428
Norm Quadratic Average: 32.30471420288086
Nearest Class Center Accuracy: 0.9415

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5119462609291077
Inter Cos: 0.3718445301055908
Norm Quadratic Average: 20.19947052001953
Nearest Class Center Accuracy: 0.9318

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6993511319160461
Inter Cos: 0.41213059425354004
Norm Quadratic Average: 15.826693534851074
Nearest Class Center Accuracy: 0.9614

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7437971234321594
Inter Cos: 0.42858442664146423
Norm Quadratic Average: 17.376502990722656
Nearest Class Center Accuracy: 0.9724

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7687134742736816
Inter Cos: 0.4313499331474304
Norm Quadratic Average: 19.610334396362305
Nearest Class Center Accuracy: 0.977

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 1.5682508945465088
Linear Weight Rank: 18
Intra Cos: 0.8335636258125305
Inter Cos: 0.35535404086112976
Norm Quadratic Average: 81.80519104003906
Nearest Class Center Accuracy: 0.9803

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 1.5756114721298218
Linear Weight Rank: 2713
Intra Cos: 0.8816884160041809
Inter Cos: 0.362055242061615
Norm Quadratic Average: 51.07732391357422
Nearest Class Center Accuracy: 0.983

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 1.5645655393600464
Linear Weight Rank: 9
Intra Cos: 0.896971583366394
Inter Cos: 0.3399359881877899
Norm Quadratic Average: 29.361034393310547
Nearest Class Center Accuracy: 0.9838

Output Layer:
Intra Cos: 0.9403575658798218
Inter Cos: 0.4043606221675873
Norm Quadratic Average: 18.64554786682129
Nearest Class Center Accuracy: 0.9861

