Model save path: ./New_Models/bn_True_dataset_MNIST_epochs_200_lr_0.01_model_type_vgg19_rand_seed_314159_test_samples_None_train_samples_None_weight_decay_0.003.pth.tar
Training Set:
Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.09116753935813904
Inter Cos: 0.10967151075601578
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: 0.058968476951122284
Inter Cos: 0.07623717933893204
Norm Quadratic Average: 2.8296070098876953
Nearest Class Center Accuracy: 0.8105

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.09981907159090042
Inter Cos: 0.09537117183208466
Norm Quadratic Average: 1.7770217657089233
Nearest Class Center Accuracy: 0.8703666666666666

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.09460999816656113
Inter Cos: 0.09043410420417786
Norm Quadratic Average: 1.404976725578308
Nearest Class Center Accuracy: 0.8756333333333334

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.16748614609241486
Inter Cos: 0.11876906454563141
Norm Quadratic Average: 1.0254545211791992
Nearest Class Center Accuracy: 0.9369333333333333

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.22424191236495972
Inter Cos: 0.13022416830062866
Norm Quadratic Average: 0.7990339994430542
Nearest Class Center Accuracy: 0.9632333333333334

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2845791280269623
Inter Cos: 0.13005223870277405
Norm Quadratic Average: 0.6655421257019043
Nearest Class Center Accuracy: 0.9751

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3292451500892639
Inter Cos: 0.12923957407474518
Norm Quadratic Average: 0.5758678913116455
Nearest Class Center Accuracy: 0.98065

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.42510342597961426
Inter Cos: 0.1292075216770172
Norm Quadratic Average: 0.41856852173805237
Nearest Class Center Accuracy: 0.9948

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6636791229248047
Inter Cos: 0.16914033889770508
Norm Quadratic Average: 0.30736830830574036
Nearest Class Center Accuracy: 0.99875

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.805377721786499
Inter Cos: 0.20468741655349731
Norm Quadratic Average: 0.25685057044029236
Nearest Class Center Accuracy: 0.9996666666666667

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8359472155570984
Inter Cos: 0.14670920372009277
Norm Quadratic Average: 0.2569982707500458
Nearest Class Center Accuracy: 0.99995

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8898996710777283
Inter Cos: 0.04945505037903786
Norm Quadratic Average: 0.29508882761001587
Nearest Class Center Accuracy: 0.9999666666666667

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9845628142356873
Inter Cos: 0.02552015706896782
Norm Quadratic Average: 0.33297085762023926
Nearest Class Center Accuracy: 1.0

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.996404767036438
Inter Cos: -0.012952709570527077
Norm Quadratic Average: 0.5685656070709229
Nearest Class Center Accuracy: 1.0

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9988126754760742
Inter Cos: -0.05326233059167862
Norm Quadratic Average: 1.0907756090164185
Nearest Class Center Accuracy: 1.0

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.1954047679901123
Linear Weight Rank: 127
Intra Cos: 0.9995132684707642
Inter Cos: -0.04189566522836685
Norm Quadratic Average: 25.8277587890625
Nearest Class Center Accuracy: 1.0

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.201981782913208
Linear Weight Rank: 1461
Intra Cos: 0.9995067715644836
Inter Cos: 0.008103916421532631
Norm Quadratic Average: 17.631376266479492
Nearest Class Center Accuracy: 1.0

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.198667287826538
Linear Weight Rank: 9
Intra Cos: 0.9995306730270386
Inter Cos: 0.03106779046356678
Norm Quadratic Average: 12.33824634552002
Nearest Class Center Accuracy: 1.0

Output Layer:
Intra Cos: 0.9996723532676697
Inter Cos: 0.05957941338419914
Norm Quadratic Average: 9.111923217773438
Nearest Class Center Accuracy: 1.0

Test Set:
Average Loss: 0.01915786392716691
Accuracy: 0.9956
NC1 Within Class Collapse: 0.08040361851453781
NC2 Equinorm: Features: 0.023125529289245605, Weights: 0.005321184638887644
NC2 Equiangle: Features: 0.060642639795939125, Weights: 0.02468730343712701
NC3 Self-Duality: 0.00826841127127409
NC4 NCC Mismatch: 9.999999999998899e-05

Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10219582915306091
Inter Cos: 0.12048851698637009
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: 0.0668289065361023
Inter Cos: 0.07888521254062653
Norm Quadratic Average: 2.818845510482788
Nearest Class Center Accuracy: 0.8213

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10972266644239426
Inter Cos: 0.09687074273824692
Norm Quadratic Average: 1.764053463935852
Nearest Class Center Accuracy: 0.8812

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1033480167388916
Inter Cos: 0.09220768511295319
Norm Quadratic Average: 1.3995957374572754
Nearest Class Center Accuracy: 0.8854

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.17861776053905487
Inter Cos: 0.11993840336799622
Norm Quadratic Average: 1.0208333730697632
Nearest Class Center Accuracy: 0.9422

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.23893031477928162
Inter Cos: 0.12693354487419128
Norm Quadratic Average: 0.7971309423446655
Nearest Class Center Accuracy: 0.9659

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.29988574981689453
Inter Cos: 0.14253097772598267
Norm Quadratic Average: 0.664486825466156
Nearest Class Center Accuracy: 0.9753

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.34455570578575134
Inter Cos: 0.1419960856437683
Norm Quadratic Average: 0.5746872425079346
Nearest Class Center Accuracy: 0.9794

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.43573808670043945
Inter Cos: 0.14049497246742249
Norm Quadratic Average: 0.41746219992637634
Nearest Class Center Accuracy: 0.9892

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.673364520072937
Inter Cos: 0.1697508692741394
Norm Quadratic Average: 0.3071783483028412
Nearest Class Center Accuracy: 0.993

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7969314455986023
Inter Cos: 0.20354875922203064
Norm Quadratic Average: 0.2568362355232239
Nearest Class Center Accuracy: 0.9945

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8381999731063843
Inter Cos: 0.1529851108789444
Norm Quadratic Average: 0.25663453340530396
Nearest Class Center Accuracy: 0.9957

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8878908753395081
Inter Cos: 0.057552993297576904
Norm Quadratic Average: 0.2942560911178589
Nearest Class Center Accuracy: 0.9956

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9734039306640625
Inter Cos: 0.025490444153547287
Norm Quadratic Average: 0.3317623436450958
Nearest Class Center Accuracy: 0.9956

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9798253178596497
Inter Cos: -0.009257848374545574
Norm Quadratic Average: 0.5663853883743286
Nearest Class Center Accuracy: 0.9956

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9809247851371765
Inter Cos: -0.0491507388651371
Norm Quadratic Average: 1.0866048336029053
Nearest Class Center Accuracy: 0.9956

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.1954047679901123
Linear Weight Rank: 127
Intra Cos: 0.9818688631057739
Inter Cos: -0.033861298114061356
Norm Quadratic Average: 25.73240089416504
Nearest Class Center Accuracy: 0.9957

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.201981782913208
Linear Weight Rank: 1461
Intra Cos: 0.9824276566505432
Inter Cos: 0.015417549759149551
Norm Quadratic Average: 17.567405700683594
Nearest Class Center Accuracy: 0.9957

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.198667287826538
Linear Weight Rank: 9
Intra Cos: 0.9828222990036011
Inter Cos: 0.040167830884456635
Norm Quadratic Average: 12.294180870056152
Nearest Class Center Accuracy: 0.9957

Output Layer:
Intra Cos: 0.9835004806518555
Inter Cos: 0.06812359392642975
Norm Quadratic Average: 9.079407691955566
Nearest Class Center Accuracy: 0.9957

