Model save path: ./New_Models/bn_False_dataset_MNIST_epochs_200_lr_0.01_model_type_vgg19_rand_seed_265358_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: 0.09116753190755844
Inter Cos: 0.10967151820659637
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: 0.11400465667247772
Inter Cos: 0.13751648366451263
Norm Quadratic Average: 70.3377456665039
Nearest Class Center Accuracy: 0.8003166666666667

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.13502287864685059
Inter Cos: 0.1750742793083191
Norm Quadratic Average: 122.96923065185547
Nearest Class Center Accuracy: 0.7849666666666667

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.14100755751132965
Inter Cos: 0.18798793852329254
Norm Quadratic Average: 227.37069702148438
Nearest Class Center Accuracy: 0.7848

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.17350715398788452
Inter Cos: 0.19700242578983307
Norm Quadratic Average: 161.13294982910156
Nearest Class Center Accuracy: 0.82

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.19552850723266602
Inter Cos: 0.22279568016529083
Norm Quadratic Average: 131.27560424804688
Nearest Class Center Accuracy: 0.8354833333333334

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.20785151422023773
Inter Cos: 0.24488143622875214
Norm Quadratic Average: 119.8182144165039
Nearest Class Center Accuracy: 0.8606666666666667

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2592145502567291
Inter Cos: 0.24982218444347382
Norm Quadratic Average: 88.92448425292969
Nearest Class Center Accuracy: 0.9122833333333333

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.24540860950946808
Inter Cos: 0.24206538498401642
Norm Quadratic Average: 31.762590408325195
Nearest Class Center Accuracy: 0.9056166666666666

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3067582845687866
Inter Cos: 0.2578940987586975
Norm Quadratic Average: 18.166160583496094
Nearest Class Center Accuracy: 0.8508166666666667

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.39165470004081726
Inter Cos: 0.26290732622146606
Norm Quadratic Average: 17.676210403442383
Nearest Class Center Accuracy: 0.8715

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5310360789299011
Inter Cos: 0.39430293440818787
Norm Quadratic Average: 20.181509017944336
Nearest Class Center Accuracy: 0.9287166666666666

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5863704681396484
Inter Cos: 0.39174187183380127
Norm Quadratic Average: 15.251880645751953
Nearest Class Center Accuracy: 0.9360666666666667

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7251662015914917
Inter Cos: 0.5192245841026306
Norm Quadratic Average: 13.698531150817871
Nearest Class Center Accuracy: 0.9629833333333333

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7740396857261658
Inter Cos: 0.4936974048614502
Norm Quadratic Average: 15.604864120483398
Nearest Class Center Accuracy: 0.9784333333333334

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.808323085308075
Inter Cos: 0.4447353184223175
Norm Quadratic Average: 18.05660629272461
Nearest Class Center Accuracy: 0.98725

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 1.688250184059143
Linear Weight Rank: 645
Intra Cos: 0.8389598727226257
Inter Cos: 0.4030900001525879
Norm Quadratic Average: 77.34029388427734
Nearest Class Center Accuracy: 0.99225

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 1.7157456874847412
Linear Weight Rank: 2654
Intra Cos: 0.8837838768959045
Inter Cos: 0.38383153080940247
Norm Quadratic Average: 50.83277893066406
Nearest Class Center Accuracy: 0.9971

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 1.693885326385498
Linear Weight Rank: 9
Intra Cos: 0.905591607093811
Inter Cos: 0.3382915258407593
Norm Quadratic Average: 31.73154067993164
Nearest Class Center Accuracy: 0.9986333333333334

Output Layer:
Intra Cos: 0.9533979892730713
Inter Cos: 0.40016692876815796
Norm Quadratic Average: 22.32131004333496
Nearest Class Center Accuracy: 0.9996

Test Set:
Average Loss: 0.04054822006432805
Accuracy: 0.9884
NC1 Within Class Collapse: 1.7183761596679688
NC2 Equinorm: Features: 0.09977428615093231, Weights: 0.04881066083908081
NC2 Equiangle: Features: 0.28878546820746526, Weights: 0.2140324698554145
NC3 Self-Duality: 0.09065653383731842
NC4 NCC Mismatch: 0.006000000000000005

Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10219582915306091
Inter Cos: 0.12048853933811188
Norm Quadratic Average: 23.595195770263672
Nearest Class Center Accuracy: 0.8229

Layer 1: Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.12554702162742615
Inter Cos: 0.15070486068725586
Norm Quadratic Average: 70.74683380126953
Nearest Class Center Accuracy: 0.8172

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1492702215909958
Inter Cos: 0.19164158403873444
Norm Quadratic Average: 123.50992584228516
Nearest Class Center Accuracy: 0.8043

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.15574240684509277
Inter Cos: 0.2061384916305542
Norm Quadratic Average: 228.3966522216797
Nearest Class Center Accuracy: 0.8045

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.18561765551567078
Inter Cos: 0.21684157848358154
Norm Quadratic Average: 161.4673309326172
Nearest Class Center Accuracy: 0.8369

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.20876988768577576
Inter Cos: 0.24492529034614563
Norm Quadratic Average: 131.494873046875
Nearest Class Center Accuracy: 0.8543

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.21966421604156494
Inter Cos: 0.2676752209663391
Norm Quadratic Average: 120.08746337890625
Nearest Class Center Accuracy: 0.8769

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2704744338989258
Inter Cos: 0.270590603351593
Norm Quadratic Average: 89.44137573242188
Nearest Class Center Accuracy: 0.9176

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2558631896972656
Inter Cos: 0.2612955570220947
Norm Quadratic Average: 31.998781204223633
Nearest Class Center Accuracy: 0.9138

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.3224342465400696
Inter Cos: 0.26287660002708435
Norm Quadratic Average: 18.288822174072266
Nearest Class Center Accuracy: 0.8639

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4116820991039276
Inter Cos: 0.270917683839798
Norm Quadratic Average: 17.794601440429688
Nearest Class Center Accuracy: 0.884

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5496792793273926
Inter Cos: 0.40334683656692505
Norm Quadratic Average: 20.337133407592773
Nearest Class Center Accuracy: 0.9301

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.595363199710846
Inter Cos: 0.38379865884780884
Norm Quadratic Average: 15.343488693237305
Nearest Class Center Accuracy: 0.9364

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7361620664596558
Inter Cos: 0.5081061124801636
Norm Quadratic Average: 13.821371078491211
Nearest Class Center Accuracy: 0.9574

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7789499163627625
Inter Cos: 0.48138707876205444
Norm Quadratic Average: 15.773354530334473
Nearest Class Center Accuracy: 0.9685

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.8087979555130005
Inter Cos: 0.4321330785751343
Norm Quadratic Average: 18.279865264892578
Nearest Class Center Accuracy: 0.976

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 1.688250184059143
Linear Weight Rank: 645
Intra Cos: 0.8384242653846741
Inter Cos: 0.3996025621891022
Norm Quadratic Average: 78.33234405517578
Nearest Class Center Accuracy: 0.9798

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 1.7157456874847412
Linear Weight Rank: 2654
Intra Cos: 0.8790650963783264
Inter Cos: 0.37988558411598206
Norm Quadratic Average: 51.53815841674805
Nearest Class Center Accuracy: 0.9845

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 1.693885326385498
Linear Weight Rank: 9
Intra Cos: 0.8981491327285767
Inter Cos: 0.33399584889411926
Norm Quadratic Average: 32.18901062011719
Nearest Class Center Accuracy: 0.9855

Output Layer:
Intra Cos: 0.9401923418045044
Inter Cos: 0.39559635519981384
Norm Quadratic Average: 22.646102905273438
Nearest Class Center Accuracy: 0.9867

