Model save path: ./New_Models/bn_False_dataset_MNIST_epochs_200_lr_0.01_model_type_vgg19_rand_seed_314159_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: 0.09116753935813904
Inter Cos: 0.10967151075601578
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: 0.1139930784702301
Inter Cos: 0.13717247545719147
Norm Quadratic Average: 68.86994171142578
Nearest Class Center Accuracy: 0.8004333333333333

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.135512113571167
Inter Cos: 0.17569005489349365
Norm Quadratic Average: 120.94218444824219
Nearest Class Center Accuracy: 0.7839166666666667

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.13951899111270905
Inter Cos: 0.18749672174453735
Norm Quadratic Average: 217.35617065429688
Nearest Class Center Accuracy: 0.7858833333333334

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.17164263129234314
Inter Cos: 0.1948869228363037
Norm Quadratic Average: 148.82568359375
Nearest Class Center Accuracy: 0.8244

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1927986443042755
Inter Cos: 0.21915887296199799
Norm Quadratic Average: 121.16221618652344
Nearest Class Center Accuracy: 0.8477666666666667

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.20520000159740448
Inter Cos: 0.23292994499206543
Norm Quadratic Average: 102.6839370727539
Nearest Class Center Accuracy: 0.8790833333333333

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.25729435682296753
Inter Cos: 0.2269318848848343
Norm Quadratic Average: 69.30874633789062
Nearest Class Center Accuracy: 0.9219166666666667

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.25534218549728394
Inter Cos: 0.2038646936416626
Norm Quadratic Average: 23.521087646484375
Nearest Class Center Accuracy: 0.90855

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.28009819984436035
Inter Cos: 0.24370348453521729
Norm Quadratic Average: 15.497389793395996
Nearest Class Center Accuracy: 0.8127166666666666

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.33712238073349
Inter Cos: 0.22214481234550476
Norm Quadratic Average: 17.433130264282227
Nearest Class Center Accuracy: 0.8209

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.4670908451080322
Inter Cos: 0.34985989332199097
Norm Quadratic Average: 21.64967155456543
Nearest Class Center Accuracy: 0.8923833333333333

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5264591574668884
Inter Cos: 0.41005876660346985
Norm Quadratic Average: 18.44240379333496
Nearest Class Center Accuracy: 0.91875

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6489573121070862
Inter Cos: 0.43685582280158997
Norm Quadratic Average: 16.452606201171875
Nearest Class Center Accuracy: 0.9604666666666667

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7263703346252441
Inter Cos: 0.44063669443130493
Norm Quadratic Average: 17.244409561157227
Nearest Class Center Accuracy: 0.9813333333333333

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7836515307426453
Inter Cos: 0.42440885305404663
Norm Quadratic Average: 18.483139038085938
Nearest Class Center Accuracy: 0.9898333333333333

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 1.6737782955169678
Linear Weight Rank: 671
Intra Cos: 0.8319011926651001
Inter Cos: 0.4092053174972534
Norm Quadratic Average: 76.54059600830078
Nearest Class Center Accuracy: 0.9936666666666667

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 1.7028337717056274
Linear Weight Rank: 2655
Intra Cos: 0.8839969635009766
Inter Cos: 0.412136971950531
Norm Quadratic Average: 49.87461853027344
Nearest Class Center Accuracy: 0.9980666666666667

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 1.6796950101852417
Linear Weight Rank: 9
Intra Cos: 0.9028006196022034
Inter Cos: 0.3797989785671234
Norm Quadratic Average: 31.42409324645996
Nearest Class Center Accuracy: 0.99895

Output Layer:
Intra Cos: 0.9446390867233276
Inter Cos: 0.4583814740180969
Norm Quadratic Average: 22.356048583984375
Nearest Class Center Accuracy: 0.9997

Test Set:
Average Loss: 0.03577133618984371
Accuracy: 0.9899
NC1 Within Class Collapse: 2.8346478939056396
NC2 Equinorm: Features: 0.097622811794281, Weights: 0.04489755630493164
NC2 Equiangle: Features: 0.265333006117079, Weights: 0.22970812055799697
NC3 Self-Duality: 0.06944877654314041
NC4 NCC Mismatch: 0.005800000000000027

Layer 0: Conv2d(1, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.10219582915306091
Inter Cos: 0.12048851698637009
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: 0.12559081614017487
Inter Cos: 0.15033645927906036
Norm Quadratic Average: 69.26673889160156
Nearest Class Center Accuracy: 0.8177

Layer 2: Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.14995045959949493
Inter Cos: 0.19240891933441162
Norm Quadratic Average: 121.47613525390625
Nearest Class Center Accuracy: 0.8035

Layer 3: Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.1544627696275711
Inter Cos: 0.20560283958911896
Norm Quadratic Average: 218.34005737304688
Nearest Class Center Accuracy: 0.8043

Layer 4: Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.18404176831245422
Inter Cos: 0.2139875292778015
Norm Quadratic Average: 149.13082885742188
Nearest Class Center Accuracy: 0.8394

Layer 5: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.20637920498847961
Inter Cos: 0.23938700556755066
Norm Quadratic Average: 121.39860534667969
Nearest Class Center Accuracy: 0.8652

Layer 6: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2173883467912674
Inter Cos: 0.25170275568962097
Norm Quadratic Average: 102.97742462158203
Nearest Class Center Accuracy: 0.8936

Layer 7: Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.26873454451560974
Inter Cos: 0.24296635389328003
Norm Quadratic Average: 69.7573471069336
Nearest Class Center Accuracy: 0.9285

Layer 8: Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2649528682231903
Inter Cos: 0.2234082818031311
Norm Quadratic Average: 23.67296600341797
Nearest Class Center Accuracy: 0.9169

Layer 9: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.2911006212234497
Inter Cos: 0.24771171808242798
Norm Quadratic Average: 15.55043888092041
Nearest Class Center Accuracy: 0.8295

Layer 10: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.34956619143486023
Inter Cos: 0.2469131499528885
Norm Quadratic Average: 17.482864379882812
Nearest Class Center Accuracy: 0.8402

Layer 11: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.47946053743362427
Inter Cos: 0.37506553530693054
Norm Quadratic Average: 21.758840560913086
Nearest Class Center Accuracy: 0.8979

Layer 12: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.5326671004295349
Inter Cos: 0.43908777832984924
Norm Quadratic Average: 18.524335861206055
Nearest Class Center Accuracy: 0.9209

Layer 13: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.6530975699424744
Inter Cos: 0.453332781791687
Norm Quadratic Average: 16.569456100463867
Nearest Class Center Accuracy: 0.9542

Layer 14: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7275634407997131
Inter Cos: 0.42281657457351685
Norm Quadratic Average: 17.407054901123047
Nearest Class Center Accuracy: 0.9726

Layer 15: Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
Intra Cos: 0.7825984358787537
Inter Cos: 0.43577301502227783
Norm Quadratic Average: 18.69158363342285
Nearest Class Center Accuracy: 0.9814

Layer 16: Linear(in_features=25088, out_features=4096, bias=True)
Linear Weight Norm: 1.6737782955169678
Linear Weight Rank: 671
Intra Cos: 0.8271008133888245
Inter Cos: 0.4215349853038788
Norm Quadratic Average: 77.49765014648438
Nearest Class Center Accuracy: 0.9843

Layer 17: Linear(in_features=4096, out_features=4096, bias=True)
Linear Weight Norm: 1.7028337717056274
Linear Weight Rank: 2655
Intra Cos: 0.8773080110549927
Inter Cos: 0.4242343008518219
Norm Quadratic Average: 50.54608917236328
Nearest Class Center Accuracy: 0.9884

Layer 18: Linear(in_features=4096, out_features=10, bias=True)
Linear Weight Norm: 1.6796950101852417
Linear Weight Rank: 9
Intra Cos: 0.8952723741531372
Inter Cos: 0.39180898666381836
Norm Quadratic Average: 31.853300094604492
Nearest Class Center Accuracy: 0.9885

Output Layer:
Intra Cos: 0.9333363175392151
Inter Cos: 0.46902531385421753
Norm Quadratic Average: 22.660520553588867
Nearest Class Center Accuracy: 0.9894

