Model save path: ./New_Models/bn_True_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: 0.058987732976675034
Inter Cos: 0.07709737122058868
Norm Quadratic Average: 2.4562835693359375
Nearest Class Center Accuracy: 0.8075166666666667

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10381057113409042
Inter Cos: 0.09940655529499054
Norm Quadratic Average: 1.4266300201416016
Nearest Class Center Accuracy: 0.8738166666666667

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10164249688386917
Inter Cos: 0.09335670620203018
Norm Quadratic Average: 1.1465388536453247
Nearest Class Center Accuracy: 0.8811833333333333

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.17634886503219604
Inter Cos: 0.11576933413743973
Norm Quadratic Average: 0.7482724785804749
Nearest Class Center Accuracy: 0.9368666666666666

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2294970154762268
Inter Cos: 0.13366730511188507
Norm Quadratic Average: 0.5539602041244507
Nearest Class Center Accuracy: 0.9613

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3016899824142456
Inter Cos: 0.13680802285671234
Norm Quadratic Average: 0.47028404474258423
Nearest Class Center Accuracy: 0.9708833333333333

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3423955738544464
Inter Cos: 0.1578395962715149
Norm Quadratic Average: 0.4185806214809418
Nearest Class Center Accuracy: 0.9759666666666666

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.42269590497016907
Inter Cos: 0.1652030199766159
Norm Quadratic Average: 0.2646934688091278
Nearest Class Center Accuracy: 0.99255

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6513308882713318
Inter Cos: 0.2739759087562561
Norm Quadratic Average: 0.17441950738430023
Nearest Class Center Accuracy: 0.9984833333333333

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.81671541929245
Inter Cos: 0.32243141531944275
Norm Quadratic Average: 0.153827503323555
Nearest Class Center Accuracy: 0.9997666666666667

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8750959634780884
Inter Cos: 0.20447134971618652
Norm Quadratic Average: 0.16904217004776
Nearest Class Center Accuracy: 0.99995

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.925519585609436
Inter Cos: 0.127738356590271
Norm Quadratic Average: 0.18563422560691833
Nearest Class Center Accuracy: 0.9999666666666667

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9864243268966675
Inter Cos: 0.14484794437885284
Norm Quadratic Average: 0.22062012553215027
Nearest Class Center Accuracy: 0.9999833333333333

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9968432784080505
Inter Cos: 0.10498946160078049
Norm Quadratic Average: 0.4688614010810852
Nearest Class Center Accuracy: 0.9999833333333333

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9986907243728638
Inter Cos: 0.15889377892017365
Norm Quadratic Average: 1.078275442123413
Nearest Class Center Accuracy: 0.9999833333333333

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.0941975116729736
Linear Weight Rank: 9
Intra Cos: 0.9991247653961182
Inter Cos: 0.2322097271680832
Norm Quadratic Average: 24.534536361694336
Nearest Class Center Accuracy: 0.9999833333333333

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.0957868099212646
Linear Weight Rank: 1548
Intra Cos: 0.999278724193573
Inter Cos: 0.22783160209655762
Norm Quadratic Average: 16.95367431640625
Nearest Class Center Accuracy: 0.9999833333333333

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.096588611602783
Linear Weight Rank: 9
Intra Cos: 0.9993187189102173
Inter Cos: 0.19780132174491882
Norm Quadratic Average: 11.945865631103516
Nearest Class Center Accuracy: 0.9999833333333333

Output Layer:
Intra Cos: 0.9994291663169861
Inter Cos: 0.11634275317192078
Norm Quadratic Average: 8.88389778137207
Nearest Class Center Accuracy: 0.9999833333333333

Test Set:
Average Loss: 0.01814467255808413
Accuracy: 0.996
NC1 Within Class Collapse: 0.09402091056108475
NC2 Equinorm: Features: 0.016121622174978256, Weights: 0.005719435401260853
NC2 Equiangle: Features: 0.1175177362230089, Weights: 0.09050496419270833
NC3 Self-Duality: 0.034480780363082886
NC4 NCC Mismatch: 0.00039999999999995595

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: 0.0671984925866127
Inter Cos: 0.07974093407392502
Norm Quadratic Average: 2.446443796157837
Nearest Class Center Accuracy: 0.8186

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1140274927020073
Inter Cos: 0.10126370936632156
Norm Quadratic Average: 1.416540265083313
Nearest Class Center Accuracy: 0.8869

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1115746721625328
Inter Cos: 0.09551432728767395
Norm Quadratic Average: 1.1422885656356812
Nearest Class Center Accuracy: 0.8907

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.18966318666934967
Inter Cos: 0.12124206125736237
Norm Quadratic Average: 0.7449040412902832
Nearest Class Center Accuracy: 0.9418

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.24613286554813385
Inter Cos: 0.13121557235717773
Norm Quadratic Average: 0.5529295206069946
Nearest Class Center Accuracy: 0.9626

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3170822262763977
Inter Cos: 0.1480049341917038
Norm Quadratic Average: 0.4697907269001007
Nearest Class Center Accuracy: 0.9708

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.35735034942626953
Inter Cos: 0.17218047380447388
Norm Quadratic Average: 0.41792309284210205
Nearest Class Center Accuracy: 0.9745

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.43670976161956787
Inter Cos: 0.17757579684257507
Norm Quadratic Average: 0.2642751932144165
Nearest Class Center Accuracy: 0.9882

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6595879197120667
Inter Cos: 0.2863667905330658
Norm Quadratic Average: 0.17435184121131897
Nearest Class Center Accuracy: 0.9928

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8140280246734619
Inter Cos: 0.32987168431282043
Norm Quadratic Average: 0.15393568575382233
Nearest Class Center Accuracy: 0.9948

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8771597146987915
Inter Cos: 0.21429097652435303
Norm Quadratic Average: 0.16881737112998962
Nearest Class Center Accuracy: 0.9956

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9214379191398621
Inter Cos: 0.13442622125148773
Norm Quadratic Average: 0.18504047393798828
Nearest Class Center Accuracy: 0.9957

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9710745811462402
Inter Cos: 0.15039530396461487
Norm Quadratic Average: 0.21984893083572388
Nearest Class Center Accuracy: 0.9957

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9760607481002808
Inter Cos: 0.11107435077428818
Norm Quadratic Average: 0.46725302934646606
Nearest Class Center Accuracy: 0.9959

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.9780580401420593
Inter Cos: 0.15787959098815918
Norm Quadratic Average: 1.0747827291488647
Nearest Class Center Accuracy: 0.9958

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 2.0941975116729736
Linear Weight Rank: 9
Intra Cos: 0.9798370003700256
Inter Cos: 0.2345293015241623
Norm Quadratic Average: 24.46299934387207
Nearest Class Center Accuracy: 0.9959

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 2.0957868099212646
Linear Weight Rank: 1548
Intra Cos: 0.9806124567985535
Inter Cos: 0.22995540499687195
Norm Quadratic Average: 16.899202346801758
Nearest Class Center Accuracy: 0.9957

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.096588611602783
Linear Weight Rank: 9
Intra Cos: 0.9810320138931274
Inter Cos: 0.20410317182540894
Norm Quadratic Average: 11.904560089111328
Nearest Class Center Accuracy: 0.9956

Output Layer:
Intra Cos: 0.9824459552764893
Inter Cos: 0.12242139875888824
Norm Quadratic Average: 8.848163604736328
Nearest Class Center Accuracy: 0.9956

