Model save path: ./New_Models/bn_False_dataset_MNIST_epochs_200_lr_0.01_model_type_vgg19_rand_seed_846264_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: 0.09116754680871964
Inter Cos: 0.10967151820659637
Norm Quadratic Average: 23.567686080932617
Nearest Class Center Accuracy: 0.8079833333333334

Layer 1: Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.11449398100376129
Inter Cos: 0.13769997656345367
Norm Quadratic Average: 64.43830871582031
Nearest Class Center Accuracy: 0.8004333333333333

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.13518893718719482
Inter Cos: 0.17493309080600739
Norm Quadratic Average: 103.65447998046875
Nearest Class Center Accuracy: 0.78545

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1419793665409088
Inter Cos: 0.18190662562847137
Norm Quadratic Average: 152.74945068359375
Nearest Class Center Accuracy: 0.80065

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.18291687965393066
Inter Cos: 0.19153299927711487
Norm Quadratic Average: 79.28083801269531
Nearest Class Center Accuracy: 0.8654833333333334

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2209775149822235
Inter Cos: 0.21559621393680573
Norm Quadratic Average: 39.97772216796875
Nearest Class Center Accuracy: 0.905

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.31419679522514343
Inter Cos: 0.2788076102733612
Norm Quadratic Average: 23.787689208984375
Nearest Class Center Accuracy: 0.9153

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3810071051120758
Inter Cos: 0.3069709837436676
Norm Quadratic Average: 23.333415985107422
Nearest Class Center Accuracy: 0.9249833333333334

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.32417428493499756
Inter Cos: 0.3184089958667755
Norm Quadratic Average: 14.635550498962402
Nearest Class Center Accuracy: 0.9054

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.34622329473495483
Inter Cos: 0.35230010747909546
Norm Quadratic Average: 12.589289665222168
Nearest Class Center Accuracy: 0.8832333333333333

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4663649797439575
Inter Cos: 0.3484068214893341
Norm Quadratic Average: 12.756453514099121
Nearest Class Center Accuracy: 0.91575

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5901079177856445
Inter Cos: 0.3881065249443054
Norm Quadratic Average: 13.965302467346191
Nearest Class Center Accuracy: 0.9601166666666666

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5782432556152344
Inter Cos: 0.4316178560256958
Norm Quadratic Average: 8.73706340789795
Nearest Class Center Accuracy: 0.9549166666666666

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7137854695320129
Inter Cos: 0.5978124737739563
Norm Quadratic Average: 8.781126022338867
Nearest Class Center Accuracy: 0.9600166666666666

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7762119770050049
Inter Cos: 0.636885941028595
Norm Quadratic Average: 11.445807456970215
Nearest Class Center Accuracy: 0.97805

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8234304189682007
Inter Cos: 0.6532309055328369
Norm Quadratic Average: 15.279932975769043
Nearest Class Center Accuracy: 0.9849

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 1.543680191040039
Linear Weight Rank: 9
Intra Cos: 0.8639130592346191
Inter Cos: 0.6304457187652588
Norm Quadratic Average: 76.81998443603516
Nearest Class Center Accuracy: 0.99035

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 1.553875207901001
Linear Weight Rank: 2580
Intra Cos: 0.8953744173049927
Inter Cos: 0.5652675032615662
Norm Quadratic Average: 55.09146499633789
Nearest Class Center Accuracy: 0.9928666666666667

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 1.5428296327590942
Linear Weight Rank: 9
Intra Cos: 0.9241699576377869
Inter Cos: 0.45637792348861694
Norm Quadratic Average: 32.14263916015625
Nearest Class Center Accuracy: 0.9950333333333333

Output Layer:
Intra Cos: 0.969838559627533
Inter Cos: 0.4674627482891083
Norm Quadratic Average: 19.939979553222656
Nearest Class Center Accuracy: 0.9967833333333334

Test Set:
Average Loss: 0.043659768315404655
Accuracy: 0.9883
NC1 Within Class Collapse: nan
NC2 Equinorm: Features: 0.12536506354808807, Weights: 0.05979420244693756
NC2 Equiangle: Features: 0.35955950419108074, Weights: 0.1751386006673177
NC3 Self-Duality: 0.19096225500106812
NC4 NCC Mismatch: 0.006800000000000028

Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10219582170248032
Inter Cos: 0.12048851698637009
Norm Quadratic Average: 23.59519386291504
Nearest Class Center Accuracy: 0.8229

Layer 1: Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.12607018649578094
Inter Cos: 0.1508825719356537
Norm Quadratic Average: 64.81484985351562
Nearest Class Center Accuracy: 0.8174

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.149520143866539
Inter Cos: 0.19143980741500854
Norm Quadratic Average: 104.1109390258789
Nearest Class Center Accuracy: 0.8048

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.15661349892616272
Inter Cos: 0.19871817529201508
Norm Quadratic Average: 153.4238739013672
Nearest Class Center Accuracy: 0.8153

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.19542300701141357
Inter Cos: 0.20820307731628418
Norm Quadratic Average: 79.48356628417969
Nearest Class Center Accuracy: 0.8763

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.23684705793857574
Inter Cos: 0.22083209455013275
Norm Quadratic Average: 40.08717727661133
Nearest Class Center Accuracy: 0.9156

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3318573236465454
Inter Cos: 0.27517905831336975
Norm Quadratic Average: 23.883481979370117
Nearest Class Center Accuracy: 0.9233

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.39768362045288086
Inter Cos: 0.30081212520599365
Norm Quadratic Average: 23.486515045166016
Nearest Class Center Accuracy: 0.9318

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.33550769090652466
Inter Cos: 0.3155747950077057
Norm Quadratic Average: 14.738030433654785
Nearest Class Center Accuracy: 0.9111

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3578398525714874
Inter Cos: 0.34695035219192505
Norm Quadratic Average: 12.689292907714844
Nearest Class Center Accuracy: 0.8877

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4842032194137573
Inter Cos: 0.34443190693855286
Norm Quadratic Average: 12.856460571289062
Nearest Class Center Accuracy: 0.9158

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6045557260513306
Inter Cos: 0.4048672616481781
Norm Quadratic Average: 14.088424682617188
Nearest Class Center Accuracy: 0.9596

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5871427059173584
Inter Cos: 0.4377789795398712
Norm Quadratic Average: 8.797228813171387
Nearest Class Center Accuracy: 0.9543

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7231247425079346
Inter Cos: 0.590938925743103
Norm Quadratic Average: 8.85191535949707
Nearest Class Center Accuracy: 0.955

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7824525833129883
Inter Cos: 0.6259912252426147
Norm Quadratic Average: 11.554898262023926
Nearest Class Center Accuracy: 0.9716

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8228890895843506
Inter Cos: 0.6487118601799011
Norm Quadratic Average: 15.441079139709473
Nearest Class Center Accuracy: 0.9777

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 1.543680191040039
Linear Weight Rank: 9
Intra Cos: 0.8624489903450012
Inter Cos: 0.6210356950759888
Norm Quadratic Average: 77.66409301757812
Nearest Class Center Accuracy: 0.9825

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 1.553875207901001
Linear Weight Rank: 2580
Intra Cos: 0.8917471766471863
Inter Cos: 0.570024847984314
Norm Quadratic Average: 55.76393127441406
Nearest Class Center Accuracy: 0.9841

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 1.5428296327590942
Linear Weight Rank: 9
Intra Cos: 0.917122483253479
Inter Cos: 0.46414434909820557
Norm Quadratic Average: 32.58269500732422
Nearest Class Center Accuracy: 0.9856

Output Layer:
Intra Cos: 0.9548471570014954
Inter Cos: 0.4611704647541046
Norm Quadratic Average: 20.227018356323242
Nearest Class Center Accuracy: 0.9867

