Model save path: ./New_Models/bn_True_dataset_MNIST_epochs_200_lr_0.01_model_type_vgg19_rand_seed_979323_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: 0.10967149585485458
Norm Quadratic Average: 23.567672729492188
Nearest Class Center Accuracy: 0.8079833333333334

Layer 1: Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.06164497882127762
Inter Cos: 0.08091037720441818
Norm Quadratic Average: 2.70001220703125
Nearest Class Center Accuracy: 0.8087833333333333

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.106386698782444
Inter Cos: 0.1001095175743103
Norm Quadratic Average: 1.5255661010742188
Nearest Class Center Accuracy: 0.8766166666666667

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10224737226963043
Inter Cos: 0.09587818384170532
Norm Quadratic Average: 1.2615931034088135
Nearest Class Center Accuracy: 0.8841

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.17871853709220886
Inter Cos: 0.11927801370620728
Norm Quadratic Average: 0.8254075646400452
Nearest Class Center Accuracy: 0.9390333333333334

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2352314442396164
Inter Cos: 0.13990649580955505
Norm Quadratic Average: 0.6181838512420654
Nearest Class Center Accuracy: 0.9621333333333333

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2974611818790436
Inter Cos: 0.15491241216659546
Norm Quadratic Average: 0.5307380557060242
Nearest Class Center Accuracy: 0.97285

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3275735676288605
Inter Cos: 0.1353892832994461
Norm Quadratic Average: 0.45992380380630493
Nearest Class Center Accuracy: 0.9768833333333333

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4111829996109009
Inter Cos: 0.12750238180160522
Norm Quadratic Average: 0.29193422198295593
Nearest Class Center Accuracy: 0.9922166666666666

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6432821750640869
Inter Cos: 0.22662246227264404
Norm Quadratic Average: 0.18790549039840698
Nearest Class Center Accuracy: 0.9980833333333333

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8238471150398254
Inter Cos: 0.30202028155326843
Norm Quadratic Average: 0.17733527719974518
Nearest Class Center Accuracy: 0.9993666666666666

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8560954928398132
Inter Cos: 0.20930615067481995
Norm Quadratic Average: 0.16447584331035614
Nearest Class Center Accuracy: 0.99995

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9100000262260437
Inter Cos: 0.17613710463047028
Norm Quadratic Average: 0.17627999186515808
Nearest Class Center Accuracy: 0.99995

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9868282079696655
Inter Cos: 0.0728982612490654
Norm Quadratic Average: 0.22742249071598053
Nearest Class Center Accuracy: 0.99995

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9972468614578247
Inter Cos: 0.06539736688137054
Norm Quadratic Average: 0.4948607087135315
Nearest Class Center Accuracy: 0.99995

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9986833930015564
Inter Cos: 0.03311600536108017
Norm Quadratic Average: 1.0951831340789795
Nearest Class Center Accuracy: 0.99995

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.1120009422302246
Linear Weight Rank: 10
Intra Cos: 0.9990694522857666
Inter Cos: 0.11506330221891403
Norm Quadratic Average: 25.13525390625
Nearest Class Center Accuracy: 0.99995

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.113758087158203
Linear Weight Rank: 1449
Intra Cos: 0.9992823004722595
Inter Cos: 0.19245845079421997
Norm Quadratic Average: 16.948366165161133
Nearest Class Center Accuracy: 0.99995

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.1144509315490723
Linear Weight Rank: 9
Intra Cos: 0.9993510246276855
Inter Cos: 0.17051197588443756
Norm Quadratic Average: 11.609569549560547
Nearest Class Center Accuracy: 0.99995

Output Layer:
Intra Cos: 0.9995353817939758
Inter Cos: 0.1470155119895935
Norm Quadratic Average: 8.399527549743652
Nearest Class Center Accuracy: 0.99995

Test Set:
Average Loss: 0.021631479407101868
Accuracy: 0.9956
NC1 Within Class Collapse: 0.0845893993973732
NC2 Equinorm: Features: 0.020844396203756332, Weights: 0.004964866675436497
NC2 Equiangle: Features: 0.0992059071858724, Weights: 0.05237721867031521
NC3 Self-Duality: 0.02399376966059208
NC4 NCC Mismatch: 0.0

Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10219583660364151
Inter Cos: 0.12048852443695068
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: 0.06996608525514603
Inter Cos: 0.08363883197307587
Norm Quadratic Average: 2.6909656524658203
Nearest Class Center Accuracy: 0.8198

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.11667446047067642
Inter Cos: 0.10163991153240204
Norm Quadratic Average: 1.5158896446228027
Nearest Class Center Accuracy: 0.8887

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.11240701377391815
Inter Cos: 0.09747350215911865
Norm Quadratic Average: 1.2580772638320923
Nearest Class Center Accuracy: 0.8933

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.19065657258033752
Inter Cos: 0.13023145496845245
Norm Quadratic Average: 0.8225725293159485
Nearest Class Center Accuracy: 0.9447

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2508583664894104
Inter Cos: 0.1518494337797165
Norm Quadratic Average: 0.6177290678024292
Nearest Class Center Accuracy: 0.9639

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3138834834098816
Inter Cos: 0.1691979020833969
Norm Quadratic Average: 0.5304433107376099
Nearest Class Center Accuracy: 0.9727

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.34378182888031006
Inter Cos: 0.1475326269865036
Norm Quadratic Average: 0.4589689075946808
Nearest Class Center Accuracy: 0.9752

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.42439067363739014
Inter Cos: 0.13976749777793884
Norm Quadratic Average: 0.29112300276756287
Nearest Class Center Accuracy: 0.9881

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6526443958282471
Inter Cos: 0.23865310847759247
Norm Quadratic Average: 0.18784992396831512
Nearest Class Center Accuracy: 0.9923

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8216590881347656
Inter Cos: 0.3110438585281372
Norm Quadratic Average: 0.17754457890987396
Nearest Class Center Accuracy: 0.9944

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8552817106246948
Inter Cos: 0.21780794858932495
Norm Quadratic Average: 0.16441380977630615
Nearest Class Center Accuracy: 0.995

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9031509757041931
Inter Cos: 0.18587501347064972
Norm Quadratic Average: 0.17591312527656555
Nearest Class Center Accuracy: 0.9951

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9719398617744446
Inter Cos: 0.07351868599653244
Norm Quadratic Average: 0.22651071846485138
Nearest Class Center Accuracy: 0.9958

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9751788973808289
Inter Cos: 0.07166063040494919
Norm Quadratic Average: 0.492811381816864
Nearest Class Center Accuracy: 0.9958

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9766781330108643
Inter Cos: 0.039467912167310715
Norm Quadratic Average: 1.0904861688613892
Nearest Class Center Accuracy: 0.9957

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.1120009422302246
Linear Weight Rank: 10
Intra Cos: 0.978401243686676
Inter Cos: 0.11859966814517975
Norm Quadratic Average: 25.036365509033203
Nearest Class Center Accuracy: 0.9957

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.113758087158203
Linear Weight Rank: 1449
Intra Cos: 0.9794277548789978
Inter Cos: 0.19368910789489746
Norm Quadratic Average: 16.880842208862305
Nearest Class Center Accuracy: 0.9956

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.1144509315490723
Linear Weight Rank: 9
Intra Cos: 0.9794519543647766
Inter Cos: 0.1723531037569046
Norm Quadratic Average: 11.562686920166016
Nearest Class Center Accuracy: 0.9956

Output Layer:
Intra Cos: 0.9802608489990234
Inter Cos: 0.1592557430267334
Norm Quadratic Average: 8.364463806152344
Nearest Class Center Accuracy: 0.9956

