Model save path: /content/drive/MyDrive/Neural Collapse/New_Models/bn_False_dataset_MNIST_epochs_100_lr_0.001_model_type_vgg11_weight_decay_0.005.pth.tar
Training Set:
Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.08946064859628677
Inter Cos: 0.11311884224414825
Norm Quadratic Average: 23.53293800354004
Nearest Class Center Accuracy: 0.81175

Layer 1: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.11466742306947708
Inter Cos: 0.14143654704093933
Norm Quadratic Average: 46.35661697387695
Nearest Class Center Accuracy: 0.810625

Layer 2: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.14894160628318787
Inter Cos: 0.17299272119998932
Norm Quadratic Average: 48.59457778930664
Nearest Class Center Accuracy: 0.783375

Layer 3: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.16071081161499023
Inter Cos: 0.19769181311130524
Norm Quadratic Average: 62.834266662597656
Nearest Class Center Accuracy: 0.78725

Layer 4: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1730942279100418
Inter Cos: 0.1885288804769516
Norm Quadratic Average: 39.37216567993164
Nearest Class Center Accuracy: 0.816625

Layer 5: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.19825239479541779
Inter Cos: 0.2177087515592575
Norm Quadratic Average: 32.72792053222656
Nearest Class Center Accuracy: 0.86675

Layer 6: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2686096727848053
Inter Cos: 0.21681050956249237
Norm Quadratic Average: 17.131370544433594
Nearest Class Center Accuracy: 0.92175

Layer 7: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4073423743247986
Inter Cos: 0.2690497040748596
Norm Quadratic Average: 11.803650856018066
Nearest Class Center Accuracy: 0.972

Layer 8: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 79.75303649902344
Linear Weight Rank: 4031
Intra Cos: 0.6427987813949585
Inter Cos: 0.27280136942863464
Norm Quadratic Average: 50.89962387084961
Nearest Class Center Accuracy: 0.99725

Layer 9: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 32.26548385620117
Linear Weight Rank: 3670
Intra Cos: 0.7512914538383484
Inter Cos: 0.2667146325111389
Norm Quadratic Average: 33.46353530883789
Nearest Class Center Accuracy: 0.999375

Layer 10: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.2610363960266113
Linear Weight Rank: 10
Intra Cos: 0.790714681148529
Inter Cos: 0.2591557502746582
Norm Quadratic Average: 26.141380310058594
Nearest Class Center Accuracy: 0.999125

Output Layer:
Intra Cos: 0.8250367641448975
Inter Cos: 0.3342002034187317
Norm Quadratic Average: 19.213640213012695
Nearest Class Center Accuracy: 0.99875

Test Set:
Average Loss: 0.07061252427101135
Accuracy: 0.979
NC1 Within Class Collapse: 2.0759899616241455
NC2 Equinorm: Features: 0.10917475819587708, Weights: 0.02151280641555786
NC2 Equiangle: Features: 0.2531663258870443, Weights: 0.1043598386976454
NC3 Self-Duality: 0.412731409072876
NC4 NCC Mismatch: 0.010499999999999954

Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10133810341358185
Inter Cos: 0.11957789957523346
Norm Quadratic Average: 22.82430076599121
Nearest Class Center Accuracy: 0.809

Layer 1: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.13727840781211853
Inter Cos: 0.15877944231033325
Norm Quadratic Average: 45.043949127197266
Nearest Class Center Accuracy: 0.806

Layer 2: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.163677379488945
Inter Cos: 0.21328580379486084
Norm Quadratic Average: 47.21666717529297
Nearest Class Center Accuracy: 0.7865

Layer 3: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.17373330891132355
Inter Cos: 0.2381390631198883
Norm Quadratic Average: 60.87213134765625
Nearest Class Center Accuracy: 0.7895

Layer 4: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.15366122126579285
Inter Cos: 0.221598282456398
Norm Quadratic Average: 38.31841278076172
Nearest Class Center Accuracy: 0.824

Layer 5: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.17526784539222717
Inter Cos: 0.24916867911815643
Norm Quadratic Average: 31.922834396362305
Nearest Class Center Accuracy: 0.862

Layer 6: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.24183140695095062
Inter Cos: 0.2388981580734253
Norm Quadratic Average: 16.659027099609375
Nearest Class Center Accuracy: 0.9115

Layer 7: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.36469006538391113
Inter Cos: 0.2936568260192871
Norm Quadratic Average: 11.436248779296875
Nearest Class Center Accuracy: 0.9495

Layer 8: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 79.75303649902344
Linear Weight Rank: 4031
Intra Cos: 0.5784250497817993
Inter Cos: 0.3059800863265991
Norm Quadratic Average: 48.96851348876953
Nearest Class Center Accuracy: 0.9695

Layer 9: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 32.26548385620117
Linear Weight Rank: 3670
Intra Cos: 0.6824563145637512
Inter Cos: 0.297171950340271
Norm Quadratic Average: 32.13650894165039
Nearest Class Center Accuracy: 0.9755

Layer 10: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 2.2610363960266113
Linear Weight Rank: 10
Intra Cos: 0.7190850377082825
Inter Cos: 0.26840364933013916
Norm Quadratic Average: 25.121826171875
Nearest Class Center Accuracy: 0.9765

Output Layer:
Intra Cos: 0.7484594583511353
Inter Cos: 0.34621816873550415
Norm Quadratic Average: 18.423351287841797
Nearest Class Center Accuracy: 0.973

