Model save path: ./New_Models/bn_True_dataset_MNIST_epochs_200_lr_0.01_model_type_vgg19_rand_seed_338327_test_samples_None_train_samples_None_weight_decay_0.0005.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: 0.10967152565717697
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: 0.06313318014144897
Inter Cos: 0.0776987299323082
Norm Quadratic Average: 20.162853240966797
Nearest Class Center Accuracy: 0.8280666666666666

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.09815644472837448
Inter Cos: 0.09476060420274734
Norm Quadratic Average: 12.012493133544922
Nearest Class Center Accuracy: 0.8724333333333333

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.09940128773450851
Inter Cos: 0.09550090879201889
Norm Quadratic Average: 12.083939552307129
Nearest Class Center Accuracy: 0.8809333333333333

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1688007116317749
Inter Cos: 0.12544508278369904
Norm Quadratic Average: 8.097912788391113
Nearest Class Center Accuracy: 0.9299

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2102433741092682
Inter Cos: 0.13702721893787384
Norm Quadratic Average: 8.634023666381836
Nearest Class Center Accuracy: 0.9534166666666667

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.24456267058849335
Inter Cos: 0.13538168370723724
Norm Quadratic Average: 8.945117950439453
Nearest Class Center Accuracy: 0.9681666666666666

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2760906219482422
Inter Cos: 0.13653752207756042
Norm Quadratic Average: 9.323158264160156
Nearest Class Center Accuracy: 0.97655

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.36537426710128784
Inter Cos: 0.10571503639221191
Norm Quadratic Average: 6.465935707092285
Nearest Class Center Accuracy: 0.9934666666666667

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5568345189094543
Inter Cos: 0.11333703994750977
Norm Quadratic Average: 7.0593767166137695
Nearest Class Center Accuracy: 0.99825

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6908915638923645
Inter Cos: 0.0833786353468895
Norm Quadratic Average: 7.580218315124512
Nearest Class Center Accuracy: 0.9995333333333334

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7951182723045349
Inter Cos: 0.03773968666791916
Norm Quadratic Average: 7.763874053955078
Nearest Class Center Accuracy: 0.9999

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8887187242507935
Inter Cos: 0.1208922415971756
Norm Quadratic Average: 6.719668388366699
Nearest Class Center Accuracy: 0.9999833333333333

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9610753059387207
Inter Cos: 0.06443887948989868
Norm Quadratic Average: 4.14648962020874
Nearest Class Center Accuracy: 1.0

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9789751768112183
Inter Cos: 0.04272991046309471
Norm Quadratic Average: 4.197505950927734
Nearest Class Center Accuracy: 1.0

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9911214113235474
Inter Cos: -0.01418857928365469
Norm Quadratic Average: 4.306966781616211
Nearest Class Center Accuracy: 1.0

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 14.675446510314941
Linear Weight Rank: 4031
Intra Cos: 0.9975120425224304
Inter Cos: -0.03638356924057007
Norm Quadratic Average: 43.193687438964844
Nearest Class Center Accuracy: 1.0

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 6.265864372253418
Linear Weight Rank: 3667
Intra Cos: 0.998201310634613
Inter Cos: 0.01910478062927723
Norm Quadratic Average: 26.984445571899414
Nearest Class Center Accuracy: 1.0

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.065049171447754
Linear Weight Rank: 10
Intra Cos: 0.997872531414032
Inter Cos: 0.04340377077460289
Norm Quadratic Average: 17.61240005493164
Nearest Class Center Accuracy: 1.0

Output Layer:
Intra Cos: 0.9996243119239807
Inter Cos: 0.11686760187149048
Norm Quadratic Average: 12.440939903259277
Nearest Class Center Accuracy: 1.0

Test Set:
Average Loss: 0.020867832660209387
Accuracy: 0.9962
NC1 Within Class Collapse: 0.0866020992398262
NC2 Equinorm: Features: 0.023728741332888603, Weights: 0.014521746896207333
NC2 Equiangle: Features: 0.09532892439100478, Weights: 0.06742242177327475
NC3 Self-Duality: 0.0475844070315361
NC4 NCC Mismatch: 0.00019999999999997797

Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10219582170248032
Inter Cos: 0.12048853188753128
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: 0.07096613198518753
Inter Cos: 0.0790383592247963
Norm Quadratic Average: 20.096973419189453
Nearest Class Center Accuracy: 0.8404

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10781550407409668
Inter Cos: 0.09543478488922119
Norm Quadratic Average: 11.941184043884277
Nearest Class Center Accuracy: 0.8831

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10890158265829086
Inter Cos: 0.09598007053136826
Norm Quadratic Average: 12.035921096801758
Nearest Class Center Accuracy: 0.8908

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1818440556526184
Inter Cos: 0.12371434271335602
Norm Quadratic Average: 8.049049377441406
Nearest Class Center Accuracy: 0.9345

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2227088212966919
Inter Cos: 0.13431449234485626
Norm Quadratic Average: 8.580936431884766
Nearest Class Center Accuracy: 0.9571

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.257631778717041
Inter Cos: 0.13181138038635254
Norm Quadratic Average: 8.898741722106934
Nearest Class Center Accuracy: 0.9689

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2893376648426056
Inter Cos: 0.13237357139587402
Norm Quadratic Average: 9.28501033782959
Nearest Class Center Accuracy: 0.9755

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3771343231201172
Inter Cos: 0.10512294620275497
Norm Quadratic Average: 6.449634075164795
Nearest Class Center Accuracy: 0.9898

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5631502270698547
Inter Cos: 0.11641775816679001
Norm Quadratic Average: 7.053103923797607
Nearest Class Center Accuracy: 0.9931

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6921308636665344
Inter Cos: 0.0905863493680954
Norm Quadratic Average: 7.577553749084473
Nearest Class Center Accuracy: 0.9944

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7932949066162109
Inter Cos: 0.045042239129543304
Norm Quadratic Average: 7.76142692565918
Nearest Class Center Accuracy: 0.9949

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8832058310508728
Inter Cos: 0.11467549949884415
Norm Quadratic Average: 6.716009616851807
Nearest Class Center Accuracy: 0.9941

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9435931444168091
Inter Cos: 0.06638386100530624
Norm Quadratic Average: 4.141315460205078
Nearest Class Center Accuracy: 0.9949

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9587244391441345
Inter Cos: 0.04314905405044556
Norm Quadratic Average: 4.189712047576904
Nearest Class Center Accuracy: 0.996

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9692177772521973
Inter Cos: -0.01397102139890194
Norm Quadratic Average: 4.297005653381348
Nearest Class Center Accuracy: 0.9961

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 14.675446510314941
Linear Weight Rank: 4031
Intra Cos: 0.9778066277503967
Inter Cos: -0.029441190883517265
Norm Quadratic Average: 43.073299407958984
Nearest Class Center Accuracy: 0.9959

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 6.265864372253418
Linear Weight Rank: 3667
Intra Cos: 0.9790581464767456
Inter Cos: 0.020539268851280212
Norm Quadratic Average: 26.908510208129883
Nearest Class Center Accuracy: 0.9962

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.065049171447754
Linear Weight Rank: 10
Intra Cos: 0.9788294434547424
Inter Cos: 0.04663228616118431
Norm Quadratic Average: 17.565153121948242
Nearest Class Center Accuracy: 0.9962

Output Layer:
Intra Cos: 0.9835206866264343
Inter Cos: 0.12248148024082184
Norm Quadratic Average: 12.403154373168945
Nearest Class Center Accuracy: 0.9962

