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_0.001.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: 0.10967151820659637
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: 0.0991009995341301
Inter Cos: 0.1207897812128067
Norm Quadratic Average: 60.104713439941406
Nearest Class Center Accuracy: 0.8082166666666667

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.12618234753608704
Inter Cos: 0.1518903225660324
Norm Quadratic Average: 74.44076538085938
Nearest Class Center Accuracy: 0.8172666666666667

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.13813434541225433
Inter Cos: 0.16778448224067688
Norm Quadratic Average: 121.06983947753906
Nearest Class Center Accuracy: 0.8218333333333333

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.20359300076961517
Inter Cos: 0.1909606158733368
Norm Quadratic Average: 101.07160949707031
Nearest Class Center Accuracy: 0.8643

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.23253192007541656
Inter Cos: 0.18752968311309814
Norm Quadratic Average: 99.97445678710938
Nearest Class Center Accuracy: 0.8873666666666666

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2492886483669281
Inter Cos: 0.1942819356918335
Norm Quadratic Average: 95.9539794921875
Nearest Class Center Accuracy: 0.9093

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.27417394518852234
Inter Cos: 0.1837453693151474
Norm Quadratic Average: 75.09016418457031
Nearest Class Center Accuracy: 0.9343666666666667

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.28426986932754517
Inter Cos: 0.1637444794178009
Norm Quadratic Average: 28.898204803466797
Nearest Class Center Accuracy: 0.9588

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4653626084327698
Inter Cos: 0.2649061977863312
Norm Quadratic Average: 15.45024299621582
Nearest Class Center Accuracy: 0.9764

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5237979888916016
Inter Cos: 0.3294418752193451
Norm Quadratic Average: 14.754018783569336
Nearest Class Center Accuracy: 0.9803833333333334

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5568668842315674
Inter Cos: 0.35639601945877075
Norm Quadratic Average: 15.965479850769043
Nearest Class Center Accuracy: 0.9858333333333333

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6225395202636719
Inter Cos: 0.3233920633792877
Norm Quadratic Average: 10.119760513305664
Nearest Class Center Accuracy: 0.9806666666666667

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7900069952011108
Inter Cos: 0.42114660143852234
Norm Quadratic Average: 8.3218994140625
Nearest Class Center Accuracy: 0.9848333333333333

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8355906009674072
Inter Cos: 0.29491928219795227
Norm Quadratic Average: 9.036808967590332
Nearest Class Center Accuracy: 0.9902

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.865176260471344
Inter Cos: 0.3122328817844391
Norm Quadratic Average: 9.536696434020996
Nearest Class Center Accuracy: 0.9927333333333334

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.902623414993286
Linear Weight Rank: 4029
Intra Cos: 0.8732001781463623
Inter Cos: 0.27463728189468384
Norm Quadratic Average: 46.500518798828125
Nearest Class Center Accuracy: 0.9940666666666667

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.2859017848968506
Linear Weight Rank: 3640
Intra Cos: 0.9080079793930054
Inter Cos: 0.31346383690834045
Norm Quadratic Average: 37.11646270751953
Nearest Class Center Accuracy: 0.9968

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.0534818172454834
Linear Weight Rank: 9
Intra Cos: 0.9180285334587097
Inter Cos: 0.3013172447681427
Norm Quadratic Average: 28.459077835083008
Nearest Class Center Accuracy: 0.9982166666666666

Output Layer:
Intra Cos: 0.9459187388420105
Inter Cos: 0.3735809624195099
Norm Quadratic Average: 24.032733917236328
Nearest Class Center Accuracy: 0.9995833333333334

Test Set:
Average Loss: 0.02505113764051348
Accuracy: 0.993
NC1 Within Class Collapse: 1.4574170112609863
NC2 Equinorm: Features: 0.1318441480398178, Weights: 0.03187286853790283
NC2 Equiangle: Features: 0.25079332987467445, Weights: 0.1915421379937066
NC3 Self-Duality: 0.11304023861885071
NC4 NCC Mismatch: 0.005199999999999982

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: 0.11054598540067673
Inter Cos: 0.13292135298252106
Norm Quadratic Average: 60.30208206176758
Nearest Class Center Accuracy: 0.821

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.14054472744464874
Inter Cos: 0.16634905338287354
Norm Quadratic Average: 74.50077819824219
Nearest Class Center Accuracy: 0.8308

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1528981626033783
Inter Cos: 0.18356654047966003
Norm Quadratic Average: 121.25516510009766
Nearest Class Center Accuracy: 0.836

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.22082193195819855
Inter Cos: 0.20844528079032898
Norm Quadratic Average: 101.08950805664062
Nearest Class Center Accuracy: 0.8773

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.25186577439308167
Inter Cos: 0.20403248071670532
Norm Quadratic Average: 99.95793914794922
Nearest Class Center Accuracy: 0.8975

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2662806808948517
Inter Cos: 0.20376022160053253
Norm Quadratic Average: 96.0038833618164
Nearest Class Center Accuracy: 0.9169

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2870907187461853
Inter Cos: 0.19711776077747345
Norm Quadratic Average: 75.29126739501953
Nearest Class Center Accuracy: 0.9426

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2961752116680145
Inter Cos: 0.1580454260110855
Norm Quadratic Average: 28.987085342407227
Nearest Class Center Accuracy: 0.9621

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.48841530084609985
Inter Cos: 0.2844737768173218
Norm Quadratic Average: 15.523015975952148
Nearest Class Center Accuracy: 0.9741

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5473105311393738
Inter Cos: 0.3547239303588867
Norm Quadratic Average: 14.855483055114746
Nearest Class Center Accuracy: 0.9771

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5782175064086914
Inter Cos: 0.3841450810432434
Norm Quadratic Average: 16.100040435791016
Nearest Class Center Accuracy: 0.9817

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6246321797370911
Inter Cos: 0.336492657661438
Norm Quadratic Average: 10.211466789245605
Nearest Class Center Accuracy: 0.9743

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7953070998191833
Inter Cos: 0.43216443061828613
Norm Quadratic Average: 8.403352737426758
Nearest Class Center Accuracy: 0.9773

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8409402966499329
Inter Cos: 0.30750852823257446
Norm Quadratic Average: 9.127572059631348
Nearest Class Center Accuracy: 0.9829

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8671209216117859
Inter Cos: 0.30836448073387146
Norm Quadratic Average: 9.630120277404785
Nearest Class Center Accuracy: 0.985

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.902623414993286
Linear Weight Rank: 4029
Intra Cos: 0.8710969686508179
Inter Cos: 0.26415035128593445
Norm Quadratic Average: 46.90214157104492
Nearest Class Center Accuracy: 0.9866

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.2859017848968506
Linear Weight Rank: 3640
Intra Cos: 0.9046922326087952
Inter Cos: 0.3019747734069824
Norm Quadratic Average: 37.43102264404297
Nearest Class Center Accuracy: 0.989

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.0534818172454834
Linear Weight Rank: 9
Intra Cos: 0.9140787124633789
Inter Cos: 0.29054754972457886
Norm Quadratic Average: 28.695335388183594
Nearest Class Center Accuracy: 0.9905

Output Layer:
Intra Cos: 0.9417651295661926
Inter Cos: 0.36245134472846985
Norm Quadratic Average: 24.231319427490234
Nearest Class Center Accuracy: 0.9923

