Model save path: /content/drive/MyDrive/Neural Collapse/New_Models/bn_False_dataset_conic_epochs_200_lr_0.001_model_depth_MLP_3_model_type_MLP_rand_seed_846264_weight_decay_0.04.pth.tar
Training Set:
Layer 0: Linear(in_features=16, out_features=200, bias=True)
Linear Weight Norm: 2.5474367141723633
Linear Weight Rank: 16
Intra Cos: 0.07988917082548141
Inter Cos: 0.0017584916204214096
Norm Quadratic Average: 4.001680850982666
Nearest Class Center Accuracy: 0.948875

Layer 1: Linear(in_features=200, out_features=200, bias=True)
Linear Weight Norm: 2.833094596862793
Linear Weight Rank: 176
Intra Cos: 0.6013579368591309
Inter Cos: 0.01421414129436016
Norm Quadratic Average: 1.8788548707962036
Nearest Class Center Accuracy: 0.9805

Layer 2: Linear(in_features=200, out_features=4, bias=True)
Linear Weight Norm: 2.687166929244995
Linear Weight Rank: 4
Intra Cos: 0.7471299171447754
Inter Cos: 0.024976054206490517
Norm Quadratic Average: 2.509390354156494
Nearest Class Center Accuracy: 0.981875

Output Layer:
Intra Cos: 0.8323890566825867
Inter Cos: 0.025895074009895325
Norm Quadratic Average: 4.518108367919922
Nearest Class Center Accuracy: 0.983375

Test Set:
Average Loss: 0.26763276958465576
Accuracy: 0.9875
NC1 Within Class Collapse: 3.426760673522949
NC2 Equinorm: Features: 0.040729280561208725, Weights: 0.016195375472307205
NC2 Equiangle: Features: 0.43812115987141925, Weights: 0.4246791998545329
NC3 Self-Duality: 0.008562829345464706
NC4 NCC Mismatch: 0.02749999999999997

Layer 0: Linear(in_features=16, out_features=200, bias=True)
Linear Weight Norm: 2.5474367141723633
Linear Weight Rank: 16
Intra Cos: 0.08033028990030289
Inter Cos: 0.003312898799777031
Norm Quadratic Average: 3.988335609436035
Nearest Class Center Accuracy: 0.9165

Layer 1: Linear(in_features=200, out_features=200, bias=True)
Linear Weight Norm: 2.833094596862793
Linear Weight Rank: 176
Intra Cos: 0.602910041809082
Inter Cos: 0.030749235302209854
Norm Quadratic Average: 1.8946764469146729
Nearest Class Center Accuracy: 0.9725

Layer 2: Linear(in_features=200, out_features=4, bias=True)
Linear Weight Norm: 2.687166929244995
Linear Weight Rank: 4
Intra Cos: 0.7418510317802429
Inter Cos: 0.04993446171283722
Norm Quadratic Average: 2.5293068885803223
Nearest Class Center Accuracy: 0.974

Output Layer:
Intra Cos: 0.8267563581466675
Inter Cos: 0.05294765904545784
Norm Quadratic Average: 4.5465922355651855
Nearest Class Center Accuracy: 0.9795

