env P{'default': None, 'device': device(type='cuda'), 'seed': 1228, 'verbose': 1, 'color': False, 'tqdm': False, 'benchmark': False, 'num_gpus': 1} ------------------------------ dataset cifar10 Parameters: CIFAR10 dataset {'data_type': 'image', 'folder_path': 'data/data/image/cifar10', 'label_names': None, 'batch_size': 96, 'num_classes': 10, 'num_workers': 4, 'valid_batch_size': 100, 'test_batch_size': 1} -------------------- imageset {'data_shape': [3, 32, 32], 'norm_par': {'mean': [0.49139968, 0.48215827, 0.44653124], 'std': [0.24703233, 0.24348505, 0.26158768]}} -------------------- ------------------------------ model resnet18_comp Parameters: ResNet model {'folder_path': 'data/model/image/cifar10'} -------------------- normalize Normalize(mean=[0.49139968, 0.48215827, 0.44653124], std=[0.24703233, 0.24348505, 0.26158768]) features Sequential pool AdaptiveAvgPool2d(output_size=(1, 1)) flatten Flatten(start_dim=1, end_dim=-1) classifier Sequential softmax Softmax(dim=1) -------------------- ------------------------------ mark mark Parameters: Watermark mark {'mark_path': 'square_white.png', 'data_shape': [3, 32, 32], 'edge_color': tensor([-1., -1., -1.]), 'mark_alpha': 0.0, 'mark_height': 3, 'mark_width': 3, 'random_pos': False, 'random_init': True, 'height_offset': 0, 'width_offset': 0} -------------------- ------------------------------ trainer trainer Parameters: Trainer optim_args {'lr': 0.025, 'parameters': 'full', 'OptimType': 'SGD', 'momentum': 0.9, 'weight_decay': 0.0003, 'lr_scheduler': False} -------------------- train_args {'epoch': 600, 'validate_interval': 1, 'verbose': 1, 'amp': False, 'save': False} -------------------- optimizer SGD ( Parameter Group 0 dampening: 0 lr: 0.025 momentum: 0.9 nesterov: False weight_decay: 0.0003 ) -------------------- ------------------------------ attack badnet Parameters: BadNet verbose {'output': [], 'indent': 0} -------------------- process {'clean_acc': 76.74, 'folder_path': 'data/attack/image/cifar10/resnet18_comp/badnet'} -------------------- badnet {'train_mode': 'batch', 'target_class': 0, 'poison_percent': 0.01, 'poison_num': 0.96} -------------------- ------------------------------ defense abs Parameters: ABS verbose {'output': [], 'indent': 0} -------------------- process {'clean_acc': 76.74, 'folder_path': 'data/defense/image/cifar10/resnet18_comp/abs'} -------------------- abs {'seed_num': 50, 'count_mask': True, 'samp_k': 1, 'same_range': False, 'n_samples': 5, 'max_troj_size': 16, 'remask_lr': 0.1, 'remask_epoch': 1000} -------------------- ------------------------------ attack results loaded from: data/attack/image/cifar10/resnet18_comp/badnet/square_white_tar0_alpha0.00_mark(3,3) Validate Clean loss: 0.258 top1: 93.400 top5: 99.760 time: 0:00:01 Validate Trigger Tgt loss: 0.000 top1: 100.000 top5: 100.000 time: 0:00:01 Validate Trigger Org loss: 9.547 top1: 10.000 top5: 92.630 time: 0:00:01 Validate Confidence: 1.000 Neuron Jaccard Idx: 0.309 sample neurons find min max label: 0 layer: features.conv1 neuron: 7 value: 0.028 layer: features.layer1.0 neuron: 7 value: 0.072 layer: features.layer1.0 neuron: 2 value: 0.434 layer: features.layer1.1 neuron: 7 value: 0.506 layer: features.layer2.0 neuron: 125 value: 0.016 layer: features.layer2.0 neuron: 71 value: 0.040 layer: features.layer2.0 neuron: 120 value: 0.102 layer: features.layer2.0 neuron: 21 value: 0.350 layer: features.layer2.0 neuron: 79 value: 0.378 layer: features.layer2.0 neuron: 118 value: 0.384 layer: features.layer2.1 neuron: 23 value: 0.075 layer: features.layer3.0 neuron: 127 value: 0.059 layer: features.layer3.0 neuron: 196 value: 0.074 layer: features.layer4.0 neuron: 420 value: 0.004 layer: features.layer4.1 neuron: 460 value: 0.000 layer: features.layer4.1 neuron: 274 value: 0.000 layer: classifier.fc neuron: 0 value: 10.057 label: 1 layer: features.layer2.1 neuron: 111 value: 0.125 layer: features.layer2.1 neuron: 127 value: 0.323 layer: features.layer3.1 neuron: 192 value: 0.012 layer: features.layer3.1 neuron: 251 value: 0.037 layer: features.layer3.1 neuron: 73 value: 0.066 layer: features.layer3.1 neuron: 52 value: 0.122 layer: features.layer4.1 neuron: 386 value: 0.000 layer: features.layer4.1 neuron: 362 value: 0.000 layer: features.layer4.1 neuron: 224 value: 0.001 layer: classifier.fc neuron: 1 value: 10.057 label: 2 layer: features.conv1 neuron: 57 value: 0.367 layer: features.layer1.0 neuron: 8 value: 0.049 layer: features.layer1.0 neuron: 5 value: 0.076 layer: features.layer1.0 neuron: 50 value: 0.098 layer: features.layer1.0 neuron: 10 value: 0.485 layer: features.layer1.0 neuron: 60 value: 0.558 layer: features.layer1.0 neuron: 61 value: 0.654 layer: features.layer1.1 neuron: 43 value: 0.032 layer: features.layer1.1 neuron: 50 value: 0.544 layer: features.layer1.1 neuron: 60 value: 2.248 layer: features.layer2.0 neuron: 25 value: 0.036 layer: features.layer2.0 neuron: 8 value: 0.278 layer: features.layer2.0 neuron: 75 value: 0.340 layer: features.layer2.0 neuron: 115 value: 0.390 layer: features.layer2.1 neuron: 115 value: 0.021 layer: features.layer2.1 neuron: 25 value: 0.098 layer: features.layer2.1 neuron: 8 value: 0.169 layer: features.layer3.0 neuron: 57 value: 0.021 layer: features.layer3.0 neuron: 56 value: 0.056 layer: features.layer3.1 neuron: 168 value: 0.002 layer: features.layer3.1 neuron: 30 value: 0.054 layer: features.layer4.0 neuron: 106 value: 0.003 layer: features.layer4.0 neuron: 21 value: 0.008 layer: features.layer4.1 neuron: 473 value: 0.001 layer: features.layer4.1 neuron: 328 value: 0.001 layer: classifier.fc neuron: 2 value: 10.057 label: 3 layer: features.conv1 neuron: 0 value: 0.069 layer: features.conv1 neuron: 11 value: 0.721 layer: features.layer1.0 neuron: 18 value: 0.060 layer: features.layer1.0 neuron: 16 value: 0.300 layer: features.layer1.0 neuron: 45 value: 0.379 layer: features.layer1.0 neuron: 32 value: 1.158 layer: features.layer1.0 neuron: 15 value: 1.298 layer: features.layer1.0 neuron: 20 value: 1.426 layer: features.layer1.1 neuron: 15 value: 0.029 layer: features.layer1.1 neuron: 42 value: 0.134 layer: features.layer1.1 neuron: 16 value: 0.214 layer: features.layer1.1 neuron: 32 value: 0.291 layer: features.layer1.1 neuron: 20 value: 0.863 layer: features.layer1.1 neuron: 37 value: 0.918 layer: features.layer1.1 neuron: 18 value: 2.750 layer: features.layer2.0 neuron: 48 value: 0.029 layer: features.layer2.0 neuron: 109 value: 0.045 layer: features.layer2.0 neuron: 0 value: 0.089 layer: features.layer2.0 neuron: 38 value: 0.106 layer: features.layer2.0 neuron: 57 value: 0.222 layer: features.layer2.0 neuron: 36 value: 0.226 layer: features.layer2.1 neuron: 16 value: 0.027 layer: features.layer2.1 neuron: 0 value: 0.091 layer: features.layer2.1 neuron: 110 value: 0.123 layer: features.layer2.1 neuron: 33 value: 0.159 layer: features.layer2.1 neuron: 20 value: 0.175 layer: features.layer2.1 neuron: 36 value: 0.178 layer: features.layer2.1 neuron: 123 value: 0.222 layer: features.layer2.1 neuron: 32 value: 0.257 layer: features.layer3.0 neuron: 147 value: 0.010 layer: features.layer3.0 neuron: 170 value: 0.021 layer: features.layer3.0 neuron: 217 value: 0.034 layer: features.layer3.0 neuron: 231 value: 0.053 layer: features.layer3.1 neuron: 83 value: 0.018 layer: features.layer3.1 neuron: 99 value: 0.031 layer: features.layer3.1 neuron: 80 value: 0.039 layer: features.layer3.1 neuron: 50 value: 0.060 layer: features.layer3.1 neuron: 147 value: 0.096 layer: features.layer4.0 neuron: 1 value: 0.003 layer: features.layer4.0 neuron: 406 value: 0.005 layer: features.layer4.1 neuron: 348 value: 0.000 layer: features.layer4.1 neuron: 350 value: 0.001 layer: classifier.fc neuron: 3 value: 10.057 label: 4 layer: features.conv1 neuron: 13 value: 0.573 layer: features.conv1 neuron: 47 value: 0.618 layer: features.layer1.0 neuron: 47 value: 0.156 layer: features.layer1.1 neuron: 36 value: 0.020 layer: features.layer1.1 neuron: 13 value: 0.027 layer: features.layer2.0 neuron: 54 value: 0.090 layer: features.layer3.0 neuron: 193 value: 0.026 layer: features.layer3.0 neuron: 4 value: 0.056 layer: features.layer3.1 neuron: 125 value: 0.004 layer: features.layer3.1 neuron: 4 value: 0.035 layer: features.layer3.1 neuron: 131 value: 0.077 layer: features.layer4.0 neuron: 465 value: 0.001 layer: features.layer4.0 neuron: 121 value: 0.001 layer: classifier.fc neuron: 4 value: 10.057 label: 5 layer: features.layer3.0 neuron: 53 value: 0.031 layer: features.layer3.1 neuron: 53 value: 0.042 layer: features.layer4.0 neuron: 453 value: 0.004 layer: classifier.fc neuron: 5 value: 10.057 label: 6 layer: features.conv1 neuron: 6 value: 0.139 layer: features.layer1.0 neuron: 49 value: 0.202 layer: features.layer1.1 neuron: 6 value: 0.075 layer: features.layer1.1 neuron: 49 value: 0.845 layer: features.layer2.0 neuron: 19 value: 0.237 layer: features.layer3.0 neuron: 74 value: 0.013 layer: features.layer3.1 neuron: 120 value: 0.016 layer: features.layer3.1 neuron: 111 value: 0.104 layer: features.layer4.0 neuron: 19 value: 0.002 layer: features.layer4.1 neuron: 439 value: 0.001 layer: features.layer4.1 neuron: 156 value: 0.001 layer: classifier.fc neuron: 6 value: 10.057 label: 7 layer: features.layer2.0 neuron: 92 value: 0.012 layer: features.layer2.1 neuron: 11 value: 0.098 layer: features.layer2.1 neuron: 65 value: 0.297 layer: features.layer2.1 neuron: 92 value: 0.322 layer: features.layer3.0 neuron: 236 value: 0.009 layer: features.layer3.0 neuron: 241 value: 0.038 layer: features.layer3.0 neuron: 252 value: 0.062 layer: features.layer3.0 neuron: 149 value: 0.065 layer: features.layer3.1 neuron: 226 value: 0.124 layer: features.layer4.0 neuron: 110 value: 0.002 layer: features.layer4.0 neuron: 4 value: 0.003 layer: features.layer4.0 neuron: 181 value: 0.004 layer: features.layer4.0 neuron: 403 value: 0.005 layer: features.layer4.1 neuron: 414 value: 0.000 layer: features.layer4.1 neuron: 57 value: 0.001 layer: classifier.fc neuron: 7 value: 10.057 label: 8 layer: features.layer2.1 neuron: 76 value: 0.086 layer: features.layer3.0 neuron: 156 value: 0.006 layer: features.layer3.0 neuron: 59 value: 0.069 layer: features.layer3.1 neuron: 177 value: 0.123 layer: features.layer4.0 neuron: 482 value: 0.002 layer: features.layer4.0 neuron: 36 value: 0.004 layer: features.layer4.0 neuron: 234 value: 0.005 layer: features.layer4.1 neuron: 20 value: 0.000 layer: features.layer4.1 neuron: 314 value: 0.000 layer: features.layer4.1 neuron: 227 value: 0.000 layer: features.layer4.1 neuron: 70 value: 0.000 layer: features.layer4.1 neuron: 52 value: 0.001 layer: classifier.fc neuron: 8 value: 10.057 label: 9 layer: features.layer2.0 neuron: 113 value: 0.272 layer: features.layer2.1 neuron: 113 value: 0.097 layer: features.layer2.1 neuron: 62 value: 0.307 layer: features.layer3.0 neuron: 122 value: 0.007 layer: features.layer3.0 neuron: 186 value: 0.025 layer: features.layer3.1 neuron: 211 value: 0.089 layer: features.layer4.0 neuron: 8 value: 0.004 layer: features.layer4.0 neuron: 286 value: 0.006 layer: features.layer4.0 neuron: 319 value: 0.008 layer: features.layer4.0 neuron: 30 value: 0.008 layer: features.layer4.1 neuron: 65 value: 0.001 layer: features.layer4.1 neuron: 205 value: 0.001 layer: classifier.fc neuron: 9 value: 10.057 remask label: 0 layer: classifier.fc neuron: 0 value: 10.057 loss: -85.731 ATK Acc: 9.520 ATK Loss: 9.374 Norm: 16.002 Score: 10.494 Jaccard: 0.000 layer: features.layer4.1 neuron: 274 value: 0.000 loss: -178.382 ATK Acc: 8.790 ATK Loss: 9.579 Norm: 16.002 Score: 10.699 Jaccard: 0.000 layer: features.layer4.1 neuron: 460 value: 0.000 loss: -197.556 ATK Acc: 5.790 ATK Loss: 8.771 Norm: 15.997 Score: 9.891 Jaccard: 0.000 layer: features.layer4.0 neuron: 420 value: 0.004 loss: -38.340 ATK Acc: 8.830 ATK Loss: 9.606 Norm: 16.001 Score: 10.726 Jaccard: 0.000 layer: features.layer3.0 neuron: 196 value: 0.074 loss: -270.218 ATK Acc: 6.140 ATK Loss: 9.645 Norm: 16.001 Score: 10.765 Jaccard: 0.000 layer: features.layer3.0 neuron: 127 value: 0.059 loss: -260.998 ATK Acc: 9.730 ATK Loss: 7.259 Norm: 24.229 Score: 8.955 Jaccard: 0.000 layer: features.layer2.1 neuron: 23 value: 0.075 loss: -6799.721 ATK Acc: 10.030 ATK Loss: 2.656 Norm: 100.088 Score: 9.662 Jaccard: 0.000 layer: features.layer2.0 neuron: 118 value: 0.384 loss: -3718.431 ATK Acc: 10.740 ATK Loss: 4.306 Norm: 100.116 Score: 11.314 Jaccard: 0.000 layer: features.layer2.0 neuron: 79 value: 0.378 loss: -5768.809 ATK Acc: 96.960 ATK Loss: 0.105 Norm: 100.729 Score: 7.156 Jaccard: 0.062 layer: features.layer2.0 neuron: 21 value: 0.350 loss: -3230.428 ATK Acc: 38.210 ATK Loss: 2.971 Norm: 100.262 Score: 9.989 Jaccard: 0.000 layer: features.layer2.0 neuron: 120 value: 0.102 loss: -5167.787 ATK Acc: 15.760 ATK Loss: 5.599 Norm: 100.911 Score: 12.662 Jaccard: 0.000 layer: features.layer2.0 neuron: 71 value: 0.040 loss: -6318.444 ATK Acc: 1.420 ATK Loss: 3.711 Norm: 101.415 Score: 10.810 Jaccard: 0.000 layer: features.layer2.0 neuron: 125 value: 0.016 loss: -4791.622 ATK Acc: 0.140 ATK Loss: 6.059 Norm: 97.209 Score: 12.864 Jaccard: 0.000 layer: features.layer1.1 neuron: 7 value: 0.506 loss: -14844.402 ATK Acc: 74.600 ATK Loss: 0.780 Norm: 101.551 Score: 7.888 Jaccard: 0.000 layer: features.layer1.0 neuron: 2 value: 0.434 loss: -8223.539 ATK Acc: 51.680 ATK Loss: 1.471 Norm: 100.245 Score: 8.488 Jaccard: 0.000 layer: features.layer1.0 neuron: 7 value: 0.072 loss: 38.426 ATK Acc: 9.070 ATK Loss: 9.638 Norm: 15.801 Score: 10.744 Jaccard: 0.000 layer: features.conv1 neuron: 7 value: 0.028 loss: 138.606 ATK Acc: 9.260 ATK Loss: 9.609 Norm: 15.998 Score: 10.729 Jaccard: 0.133 Label: 0 loss: -5768.809 ATK Acc: 96.960 ATK loss: 0.105 Norm: 100.729 Jaccard: 0.062 Score: 7.156 label: 1 layer: classifier.fc neuron: 1 value: 10.057 loss: -57.899 ATK Acc: 7.760 ATK Loss: 9.683 Norm: 16.000 Score: 10.803 Jaccard: 0.000 layer: features.layer4.1 neuron: 224 value: 0.001 loss: -158.929 ATK Acc: 6.830 ATK Loss: 9.812 Norm: 15.998 Score: 10.932 Jaccard: 0.000 layer: features.layer4.1 neuron: 362 value: 0.000 loss: -266.873 ATK Acc: 7.830 ATK Loss: 9.610 Norm: 15.999 Score: 10.730 Jaccard: 0.000 layer: features.layer4.1 neuron: 386 value: 0.000 loss: -326.353 ATK Acc: 8.060 ATK Loss: 9.723 Norm: 16.001 Score: 10.843 Jaccard: 0.000 layer: features.layer3.1 neuron: 52 value: 0.122 loss: -150.189 ATK Acc: 8.950 ATK Loss: 9.599 Norm: 16.002 Score: 10.719 Jaccard: 0.000 layer: features.layer3.1 neuron: 73 value: 0.066 loss: -807.543 ATK Acc: 1.980 ATK Loss: 8.093 Norm: 44.521 Score: 11.209 Jaccard: 0.000 layer: features.layer3.1 neuron: 251 value: 0.037 loss: -256.701 ATK Acc: 12.300 ATK Loss: 7.950 Norm: 21.539 Score: 9.458 Jaccard: 0.000 layer: features.layer3.1 neuron: 192 value: 0.012 loss: -182.602 ATK Acc: 8.180 ATK Loss: 9.850 Norm: 16.001 Score: 10.970 Jaccard: 0.000 layer: features.layer2.1 neuron: 127 value: 0.323 loss: -5539.839 ATK Acc: 1.780 ATK Loss: 7.964 Norm: 101.244 Score: 15.051 Jaccard: 0.000 layer: features.layer2.1 neuron: 111 value: 0.125 loss: -2833.803 ATK Acc: 3.700 ATK Loss: 6.148 Norm: 100.412 Score: 13.177 Jaccard: 0.000 Label: 1 loss: -256.701 ATK Acc: 12.300 ATK loss: 7.950 Norm: 21.539 Jaccard: 0.000 Score: 9.458 label: 2 layer: classifier.fc neuron: 2 value: 10.057 loss: -159.065 ATK Acc: 2.510 ATK Loss: 9.063 Norm: 19.039 Score: 10.396 Jaccard: 0.000 layer: features.layer4.1 neuron: 328 value: 0.001 loss: -167.347 ATK Acc: 5.590 ATK Loss: 9.577 Norm: 15.999 Score: 10.697 Jaccard: 0.000 layer: features.layer4.1 neuron: 473 value: 0.001 loss: -659.938 ATK Acc: 0.190 ATK Loss: 8.903 Norm: 32.142 Score: 11.153 Jaccard: 0.000 layer: features.layer4.0 neuron: 21 value: 0.008 loss: -8.189 ATK Acc: 8.060 ATK Loss: 9.781 Norm: 15.999 Score: 10.901 Jaccard: 0.000 layer: features.layer4.0 neuron: 106 value: 0.003 loss: -24.011 ATK Acc: 7.700 ATK Loss: 9.584 Norm: 15.997 Score: 10.703 Jaccard: 0.000 layer: features.layer3.1 neuron: 30 value: 0.054 loss: -161.769 ATK Acc: 8.640 ATK Loss: 9.715 Norm: 16.001 Score: 10.835 Jaccard: 0.000 layer: features.layer3.1 neuron: 168 value: 0.002 loss: -1074.754 ATK Acc: 0.280 ATK Loss: 11.766 Norm: 59.307 Score: 15.917 Jaccard: 0.000 layer: features.layer3.0 neuron: 56 value: 0.056 loss: -998.239 ATK Acc: 0.050 ATK Loss: 11.787 Norm: 59.218 Score: 15.932 Jaccard: 0.000 layer: features.layer3.0 neuron: 57 value: 0.021 loss: -401.835 ATK Acc: 0.100 ATK Loss: 6.467 Norm: 35.667 Score: 8.964 Jaccard: 0.000 layer: features.layer2.1 neuron: 8 value: 0.169 loss: -2183.691 ATK Acc: 83.200 ATK Loss: 0.600 Norm: 56.341 Score: 4.544 Jaccard: 0.133 layer: features.layer2.1 neuron: 25 value: 0.098 loss: -4858.809 ATK Acc: 0.020 ATK Loss: 11.302 Norm: 100.524 Score: 18.339 Jaccard: 0.000 layer: features.layer2.1 neuron: 115 value: 0.021 loss: -9327.226 ATK Acc: 1.430 ATK Loss: 4.242 Norm: 100.816 Score: 11.299 Jaccard: 0.000 layer: features.layer2.0 neuron: 115 value: 0.390 loss: -9210.386 ATK Acc: 41.100 ATK Loss: 1.356 Norm: 100.650 Score: 8.402 Jaccard: 0.062 layer: features.layer2.0 neuron: 75 value: 0.340 loss: -10161.954 ATK Acc: 0.000 ATK Loss: 11.369 Norm: 101.443 Score: 18.470 Jaccard: 0.000 layer: features.layer2.0 neuron: 8 value: 0.278 loss: -1654.886 ATK Acc: 28.610 ATK Loss: 4.789 Norm: 30.997 Score: 6.958 Jaccard: 0.133 layer: features.layer2.0 neuron: 25 value: 0.036 loss: -6557.287 ATK Acc: 0.010 ATK Loss: 10.970 Norm: 100.188 Score: 17.984 Jaccard: 0.000 layer: features.layer1.1 neuron: 60 value: 2.248 loss: -11744.139 ATK Acc: 0.570 ATK Loss: 7.344 Norm: 102.168 Score: 14.496 Jaccard: 0.000 layer: features.layer1.1 neuron: 50 value: 0.544 loss: -12077.045 ATK Acc: 0.290 ATK Loss: 9.160 Norm: 101.825 Score: 16.288 Jaccard: 0.000 layer: features.layer1.1 neuron: 43 value: 0.032 loss: -24654.684 ATK Acc: 30.670 ATK Loss: 1.946 Norm: 100.306 Score: 8.967 Jaccard: 0.000 layer: features.layer1.0 neuron: 61 value: 0.654 loss: -30295.922 ATK Acc: 0.000 ATK Loss: 9.498 Norm: 101.698 Score: 16.616 Jaccard: 0.000 layer: features.layer1.0 neuron: 60 value: 0.558 loss: -7336.018 ATK Acc: 34.260 ATK Loss: 2.292 Norm: 100.404 Score: 9.320 Jaccard: 0.062 layer: features.layer1.0 neuron: 10 value: 0.485 loss: -50411.109 ATK Acc: 1.230 ATK Loss: 3.021 Norm: 101.465 Score: 10.124 Jaccard: 0.000 layer: features.layer1.0 neuron: 50 value: 0.098 loss: -13257.748 ATK Acc: 5.550 ATK Loss: 7.549 Norm: 100.437 Score: 14.579 Jaccard: 0.000 layer: features.layer1.0 neuron: 5 value: 0.076 loss: -18885.367 ATK Acc: 4.750 ATK Loss: 6.111 Norm: 100.863 Score: 13.171 Jaccard: 0.062 layer: features.layer1.0 neuron: 8 value: 0.049 loss: -13935.062 ATK Acc: 10.230 ATK Loss: 4.786 Norm: 100.063 Score: 11.790 Jaccard: 0.000 layer: features.conv1 neuron: 57 value: 0.367 loss: -14160.925 ATK Acc: 11.670 ATK Loss: 5.798 Norm: 100.017 Score: 12.799 Jaccard: 0.000 Label: 2 loss: -2183.691 ATK Acc: 83.200 ATK loss: 0.600 Norm: 56.341 Jaccard: 0.133 Score: 4.544 label: 3 layer: classifier.fc neuron: 3 value: 10.057 loss: -170.226 ATK Acc: 4.280 ATK Loss: 9.702 Norm: 16.001 Score: 10.822 Jaccard: 0.000 layer: features.layer4.1 neuron: 350 value: 0.001 loss: -157.639 ATK Acc: 8.570 ATK Loss: 9.538 Norm: 16.001 Score: 10.658 Jaccard: 0.000 layer: features.layer4.1 neuron: 348 value: 0.000 loss: -191.491 ATK Acc: 8.930 ATK Loss: 9.655 Norm: 15.997 Score: 10.775 Jaccard: 0.000 layer: features.layer4.0 neuron: 406 value: 0.005 loss: -22.469 ATK Acc: 6.870 ATK Loss: 9.835 Norm: 16.000 Score: 10.955 Jaccard: 0.000 layer: features.layer4.0 neuron: 1 value: 0.003 loss: -14.029 ATK Acc: 8.090 ATK Loss: 9.664 Norm: 15.998 Score: 10.784 Jaccard: 0.000 layer: features.layer3.1 neuron: 147 value: 0.096 loss: -175.388 ATK Acc: 8.980 ATK Loss: 9.400 Norm: 15.997 Score: 10.520 Jaccard: 0.000 layer: features.layer3.1 neuron: 50 value: 0.060 loss: -158.418 ATK Acc: 7.650 ATK Loss: 9.654 Norm: 16.001 Score: 10.774 Jaccard: 0.000 layer: features.layer3.1 neuron: 80 value: 0.039 loss: -387.919 ATK Acc: 6.590 ATK Loss: 9.918 Norm: 16.002 Score: 11.038 Jaccard: 0.000 layer: features.layer3.1 neuron: 99 value: 0.031 loss: -158.698 ATK Acc: 8.490 ATK Loss: 9.710 Norm: 16.001 Score: 10.830 Jaccard: 0.000 layer: features.layer3.1 neuron: 83 value: 0.018 loss: -289.971 ATK Acc: 5.470 ATK Loss: 9.431 Norm: 21.797 Score: 10.956 Jaccard: 0.000 layer: features.layer3.0 neuron: 231 value: 0.053 loss: -278.515 ATK Acc: 8.920 ATK Loss: 9.681 Norm: 16.001 Score: 10.801 Jaccard: 0.000 layer: features.layer3.0 neuron: 217 value: 0.034 loss: -569.744 ATK Acc: 0.930 ATK Loss: 9.707 Norm: 24.695 Score: 11.435 Jaccard: 0.000 layer: features.layer3.0 neuron: 170 value: 0.021 loss: -518.869 ATK Acc: 0.010 ATK Loss: 7.484 Norm: 44.710 Score: 10.614 Jaccard: 0.000 layer: features.layer3.0 neuron: 147 value: 0.010 loss: -184.798 ATK Acc: 8.700 ATK Loss: 9.535 Norm: 15.999 Score: 10.655 Jaccard: 0.000 layer: features.layer2.1 neuron: 32 value: 0.257 loss: -7056.414 ATK Acc: 0.000 ATK Loss: 9.810 Norm: 100.930 Score: 16.875 Jaccard: 0.000 layer: features.layer2.1 neuron: 123 value: 0.222 loss: -4728.758 ATK Acc: 43.920 ATK Loss: 2.337 Norm: 100.337 Score: 9.361 Jaccard: 0.000 layer: features.layer2.1 neuron: 36 value: 0.178 loss: -5293.091 ATK Acc: 16.290 ATK Loss: 2.426 Norm: 100.671 Score: 9.473 Jaccard: 0.062 layer: features.layer2.1 neuron: 20 value: 0.175 loss: -12375.082 ATK Acc: 0.000 ATK Loss: 7.892 Norm: 100.865 Score: 14.953 Jaccard: 0.000 layer: features.layer2.1 neuron: 33 value: 0.159 loss: -6608.153 ATK Acc: 0.000 ATK Loss: 14.157 Norm: 101.264 Score: 21.245 Jaccard: 0.000 layer: features.layer2.1 neuron: 110 value: 0.123 loss: -5786.837 ATK Acc: 4.050 ATK Loss: 5.849 Norm: 101.760 Score: 12.972 Jaccard: 0.000 layer: features.layer2.1 neuron: 0 value: 0.091 loss: -6141.115 ATK Acc: 0.010 ATK Loss: 8.482 Norm: 101.458 Score: 15.584 Jaccard: 0.000 layer: features.layer2.1 neuron: 16 value: 0.027 loss: -3094.258 ATK Acc: 0.010 ATK Loss: 7.173 Norm: 100.053 Score: 14.177 Jaccard: 0.000 layer: features.layer2.0 neuron: 36 value: 0.226 loss: -3299.941 ATK Acc: 0.010 ATK Loss: 8.386 Norm: 96.334 Score: 15.129 Jaccard: 0.000 layer: features.layer2.0 neuron: 57 value: 0.222 loss: -3697.691 ATK Acc: 28.520 ATK Loss: 2.746 Norm: 100.012 Score: 9.747 Jaccard: 0.062 layer: features.layer2.0 neuron: 38 value: 0.106 loss: -6937.291 ATK Acc: 96.970 ATK Loss: 0.131 Norm: 101.225 Score: 7.217 Jaccard: 0.062 layer: features.layer2.0 neuron: 0 value: 0.089 loss: -6596.315 ATK Acc: 0.000 ATK Loss: 8.964 Norm: 101.703 Score: 16.083 Jaccard: 0.000 layer: features.layer2.0 neuron: 109 value: 0.045 loss: -815.514 ATK Acc: 9.190 ATK Loss: 9.606 Norm: 16.002 Score: 10.727 Jaccard: 0.000 layer: features.layer2.0 neuron: 48 value: 0.029 loss: -2293.425 ATK Acc: 0.450 ATK Loss: 7.593 Norm: 82.916 Score: 13.397 Jaccard: 0.000 layer: features.layer1.1 neuron: 18 value: 2.750 loss: -40522.012 ATK Acc: 0.000 ATK Loss: 9.803 Norm: 99.824 Score: 16.791 Jaccard: 0.062 layer: features.layer1.1 neuron: 37 value: 0.918 loss: -50247.531 ATK Acc: 0.680 ATK Loss: 5.556 Norm: 100.228 Score: 12.572 Jaccard: 0.000 layer: features.layer1.1 neuron: 20 value: 0.863 loss: -26812.316 ATK Acc: 0.050 ATK Loss: 8.904 Norm: 99.352 Score: 15.859 Jaccard: 0.000 layer: features.layer1.1 neuron: 32 value: 0.291 loss: -17652.887 ATK Acc: 2.040 ATK Loss: 6.116 Norm: 100.303 Score: 13.138 Jaccard: 0.000 layer: features.layer1.1 neuron: 16 value: 0.214 loss: -15856.726 ATK Acc: 2.660 ATK Loss: 5.582 Norm: 101.521 Score: 12.689 Jaccard: 0.000 layer: features.layer1.1 neuron: 42 value: 0.134 loss: -62822.500 ATK Acc: 0.010 ATK Loss: 5.123 Norm: 101.705 Score: 12.242 Jaccard: 0.000 layer: features.layer1.1 neuron: 15 value: 0.029 loss: -21262.314 ATK Acc: 0.050 ATK Loss: 5.825 Norm: 100.040 Score: 12.828 Jaccard: 0.062 layer: features.layer1.0 neuron: 20 value: 1.426 loss: -23763.160 ATK Acc: 0.030 ATK Loss: 6.275 Norm: 101.599 Score: 13.387 Jaccard: 0.062 layer: features.layer1.0 neuron: 15 value: 1.298 loss: -8889.011 ATK Acc: 12.740 ATK Loss: 4.750 Norm: 100.233 Score: 11.766 Jaccard: 0.000 layer: features.layer1.0 neuron: 32 value: 1.158 loss: -4224.354 ATK Acc: 9.360 ATK Loss: 8.615 Norm: 59.688 Score: 12.793 Jaccard: 0.000 layer: features.layer1.0 neuron: 45 value: 0.379 loss: -9364.272 ATK Acc: 0.270 ATK Loss: 8.494 Norm: 100.289 Score: 15.514 Jaccard: 0.000 layer: features.layer1.0 neuron: 16 value: 0.300 loss: -15551.857 ATK Acc: 19.510 ATK Loss: 3.352 Norm: 101.949 Score: 10.489 Jaccard: 0.000 layer: features.layer1.0 neuron: 18 value: 0.060 loss: -16005.122 ATK Acc: 10.310 ATK Loss: 2.387 Norm: 101.621 Score: 9.501 Jaccard: 0.062 layer: features.conv1 neuron: 11 value: 0.721 loss: -303.789 ATK Acc: 8.710 ATK Loss: 9.415 Norm: 100.427 Score: 16.445 Jaccard: 0.000 layer: features.conv1 neuron: 0 value: 0.069 loss: -3517.261 ATK Acc: 8.930 ATK Loss: 9.084 Norm: 100.357 Score: 16.109 Jaccard: 0.000 Label: 3 loss: -6937.291 ATK Acc: 96.970 ATK loss: 0.131 Norm: 101.225 Jaccard: 0.062 Score: 7.217 label: 4 layer: classifier.fc neuron: 4 value: 10.057 loss: -132.482 ATK Acc: 5.330 ATK Loss: 9.887 Norm: 16.004 Score: 11.007 Jaccard: 0.000 layer: features.layer4.0 neuron: 121 value: 0.001 loss: -12.672 ATK Acc: 8.990 ATK Loss: 9.662 Norm: 15.995 Score: 10.782 Jaccard: 0.000 layer: features.layer4.0 neuron: 465 value: 0.001 loss: -20.508 ATK Acc: 8.270 ATK Loss: 9.551 Norm: 16.000 Score: 10.671 Jaccard: 0.000 layer: features.layer3.1 neuron: 131 value: 0.077 loss: -2037.938 ATK Acc: 63.480 ATK Loss: 0.974 Norm: 70.141 Score: 5.883 Jaccard: 0.000 layer: features.layer3.1 neuron: 4 value: 0.035 loss: -201.220 ATK Acc: 8.620 ATK Loss: 9.534 Norm: 16.002 Score: 10.654 Jaccard: 0.000 layer: features.layer3.1 neuron: 125 value: 0.004 loss: -359.177 ATK Acc: 7.350 ATK Loss: 9.143 Norm: 34.677 Score: 11.570 Jaccard: 0.000 layer: features.layer3.0 neuron: 4 value: 0.056 loss: -199.713 ATK Acc: 8.540 ATK Loss: 9.541 Norm: 16.001 Score: 10.661 Jaccard: 0.000 layer: features.layer3.0 neuron: 193 value: 0.026 loss: -1870.780 ATK Acc: 0.010 ATK Loss: 5.792 Norm: 100.058 Score: 12.796 Jaccard: 0.000 layer: features.layer2.0 neuron: 54 value: 0.090 loss: -6605.849 ATK Acc: 7.650 ATK Loss: 3.484 Norm: 101.977 Score: 10.622 Jaccard: 0.000 layer: features.layer1.1 neuron: 13 value: 0.027 loss: -11432.926 ATK Acc: 0.350 ATK Loss: 7.765 Norm: 101.299 Score: 14.856 Jaccard: 0.000 layer: features.layer1.1 neuron: 36 value: 0.020 loss: -10388.411 ATK Acc: 9.190 ATK Loss: 9.621 Norm: 16.001 Score: 10.741 Jaccard: 0.000 layer: features.layer1.0 neuron: 47 value: 0.156 loss: -33472.031 ATK Acc: 1.680 ATK Loss: 3.018 Norm: 101.909 Score: 10.152 Jaccard: 0.000 layer: features.conv1 neuron: 47 value: 0.618 loss: -10696.366 ATK Acc: 9.040 ATK Loss: 9.076 Norm: 100.515 Score: 16.112 Jaccard: 0.000 layer: features.conv1 neuron: 13 value: 0.573 loss: -3253.307 ATK Acc: 9.320 ATK Loss: 9.392 Norm: 79.819 Score: 14.979 Jaccard: 0.000 Label: 4 loss: -2037.938 ATK Acc: 63.480 ATK loss: 0.974 Norm: 70.141 Jaccard: 0.000 Score: 5.883 label: 5 layer: classifier.fc neuron: 5 value: 10.057 loss: -108.816 ATK Acc: 8.360 ATK Loss: 9.760 Norm: 16.001 Score: 10.881 Jaccard: 0.000 layer: features.layer4.0 neuron: 453 value: 0.004 loss: -22.174 ATK Acc: 5.570 ATK Loss: 9.775 Norm: 15.990 Score: 10.894 Jaccard: 0.000 layer: features.layer3.1 neuron: 53 value: 0.042 loss: -143.008 ATK Acc: 7.430 ATK Loss: 9.508 Norm: 16.000 Score: 10.628 Jaccard: 0.214 layer: features.layer3.0 neuron: 53 value: 0.031 loss: -210.272 ATK Acc: 6.430 ATK Loss: 8.908 Norm: 16.000 Score: 10.028 Jaccard: 0.133 Label: 5 loss: -210.272 ATK Acc: 6.430 ATK loss: 8.908 Norm: 16.000 Jaccard: 0.133 Score: 10.028 label: 6 layer: classifier.fc neuron: 6 value: 10.057 loss: -87.625 ATK Acc: 7.540 ATK Loss: 9.715 Norm: 16.001 Score: 10.835 Jaccard: 0.000 layer: features.layer4.1 neuron: 156 value: 0.001 loss: -235.866 ATK Acc: 7.990 ATK Loss: 9.677 Norm: 16.003 Score: 10.797 Jaccard: 0.000 layer: features.layer4.1 neuron: 439 value: 0.001 loss: -430.957 ATK Acc: 1.220 ATK Loss: 9.584 Norm: 31.030 Score: 11.756 Jaccard: 0.000 layer: features.layer4.0 neuron: 19 value: 0.002 loss: -21.791 ATK Acc: 8.980 ATK Loss: 9.644 Norm: 15.998 Score: 10.764 Jaccard: 0.000 layer: features.layer3.1 neuron: 111 value: 0.104 loss: -849.459 ATK Acc: 2.310 ATK Loss: 8.427 Norm: 53.433 Score: 12.167 Jaccard: 0.000 layer: features.layer3.1 neuron: 120 value: 0.016 loss: -1099.487 ATK Acc: 0.820 ATK Loss: 7.801 Norm: 50.844 Score: 11.360 Jaccard: 0.000 layer: features.layer3.0 neuron: 74 value: 0.013 loss: -227.721 ATK Acc: 4.570 ATK Loss: 9.650 Norm: 16.001 Score: 10.770 Jaccard: 0.000 layer: features.layer2.0 neuron: 19 value: 0.237 loss: -5121.201 ATK Acc: 0.020 ATK Loss: 10.245 Norm: 100.846 Score: 17.305 Jaccard: 0.000 layer: features.layer1.1 neuron: 49 value: 0.845 loss: -37383.695 ATK Acc: 0.000 ATK Loss: 13.173 Norm: 101.126 Score: 20.251 Jaccard: 0.000 layer: features.layer1.1 neuron: 6 value: 0.075 loss: -42793.242 ATK Acc: 0.010 ATK Loss: 9.023 Norm: 100.655 Score: 16.069 Jaccard: 0.000 layer: features.layer1.0 neuron: 49 value: 0.202 loss: -17516.770 ATK Acc: 0.010 ATK Loss: 10.970 Norm: 100.497 Score: 18.005 Jaccard: 0.000 layer: features.conv1 neuron: 6 value: 0.139 loss: -32018.492 ATK Acc: 3.570 ATK Loss: 8.691 Norm: 101.105 Score: 15.768 Jaccard: 0.062 Label: 6 loss: -21.791 ATK Acc: 8.980 ATK loss: 9.644 Norm: 15.998 Jaccard: 0.000 Score: 10.764 label: 7 layer: classifier.fc neuron: 7 value: 10.057 loss: -62.631 ATK Acc: 8.600 ATK Loss: 9.694 Norm: 16.001 Score: 10.814 Jaccard: 0.000 layer: features.layer4.1 neuron: 57 value: 0.001 loss: -311.303 ATK Acc: 8.020 ATK Loss: 9.717 Norm: 16.001 Score: 10.837 Jaccard: 0.000 layer: features.layer4.1 neuron: 414 value: 0.000 loss: -295.836 ATK Acc: 7.580 ATK Loss: 9.752 Norm: 16.001 Score: 10.872 Jaccard: 0.000 layer: features.layer4.0 neuron: 403 value: 0.005 loss: -24.161 ATK Acc: 8.850 ATK Loss: 9.567 Norm: 15.997 Score: 10.687 Jaccard: 0.000 layer: features.layer4.0 neuron: 181 value: 0.004 loss: -20.881 ATK Acc: 8.210 ATK Loss: 9.781 Norm: 15.996 Score: 10.901 Jaccard: 0.000 layer: features.layer4.0 neuron: 4 value: 0.003 loss: -18.566 ATK Acc: 6.600 ATK Loss: 9.848 Norm: 16.000 Score: 10.968 Jaccard: 0.000 layer: features.layer4.0 neuron: 110 value: 0.002 loss: -8.168 ATK Acc: 8.440 ATK Loss: 9.725 Norm: 15.993 Score: 10.845 Jaccard: 0.214 layer: features.layer3.1 neuron: 226 value: 0.124 loss: -164.912 ATK Acc: 7.420 ATK Loss: 9.601 Norm: 16.001 Score: 10.721 Jaccard: 0.000 layer: features.layer3.0 neuron: 149 value: 0.065 loss: -254.510 ATK Acc: 9.160 ATK Loss: 9.677 Norm: 16.000 Score: 10.797 Jaccard: 0.000 layer: features.layer3.0 neuron: 252 value: 0.062 loss: -188.254 ATK Acc: 6.160 ATK Loss: 9.619 Norm: 19.171 Score: 10.961 Jaccard: 0.133 layer: features.layer3.0 neuron: 241 value: 0.038 loss: -1182.854 ATK Acc: 0.000 ATK Loss: 10.808 Norm: 75.835 Score: 16.116 Jaccard: 0.000 layer: features.layer3.0 neuron: 236 value: 0.009 loss: -72.914 ATK Acc: 7.360 ATK Loss: 9.826 Norm: 15.996 Score: 10.946 Jaccard: 0.000 layer: features.layer2.1 neuron: 92 value: 0.322 loss: -8034.789 ATK Acc: 37.570 ATK Loss: 2.412 Norm: 99.955 Score: 9.409 Jaccard: 0.000 layer: features.layer2.1 neuron: 65 value: 0.297 loss: -6262.049 ATK Acc: 26.450 ATK Loss: 2.660 Norm: 100.810 Score: 9.716 Jaccard: 0.000 layer: features.layer2.1 neuron: 11 value: 0.098 loss: -13030.828 ATK Acc: 0.000 ATK Loss: 7.121 Norm: 101.006 Score: 14.192 Jaccard: 0.000 layer: features.layer2.0 neuron: 92 value: 0.012 loss: -6576.585 ATK Acc: 0.540 ATK Loss: 6.962 Norm: 100.410 Score: 13.991 Jaccard: 0.000 Label: 7 loss: -8034.789 ATK Acc: 37.570 ATK loss: 2.412 Norm: 99.955 Jaccard: 0.000 Score: 9.409 label: 8 layer: classifier.fc neuron: 8 value: 10.057 loss: -104.005 ATK Acc: 5.620 ATK Loss: 9.690 Norm: 16.002 Score: 10.810 Jaccard: 0.000 layer: features.layer4.1 neuron: 52 value: 0.001 loss: -296.797 ATK Acc: 8.360 ATK Loss: 9.359 Norm: 16.000 Score: 10.479 Jaccard: 0.000 layer: features.layer4.1 neuron: 70 value: 0.000 loss: -260.049 ATK Acc: 6.770 ATK Loss: 9.715 Norm: 16.001 Score: 10.835 Jaccard: 0.000 layer: features.layer4.1 neuron: 227 value: 0.000 loss: -193.250 ATK Acc: 4.020 ATK Loss: 9.536 Norm: 16.001 Score: 10.656 Jaccard: 0.000 layer: features.layer4.1 neuron: 314 value: 0.000 loss: -316.082 ATK Acc: 7.700 ATK Loss: 8.987 Norm: 16.002 Score: 10.108 Jaccard: 0.000 layer: features.layer4.1 neuron: 20 value: 0.000 loss: -241.767 ATK Acc: 5.660 ATK Loss: 9.875 Norm: 16.002 Score: 10.996 Jaccard: 0.000 layer: features.layer4.0 neuron: 234 value: 0.005 loss: -12.245 ATK Acc: 4.600 ATK Loss: 9.703 Norm: 15.997 Score: 10.823 Jaccard: 0.000 layer: features.layer4.0 neuron: 36 value: 0.004 loss: -15.996 ATK Acc: 8.300 ATK Loss: 9.710 Norm: 15.997 Score: 10.830 Jaccard: 0.000 layer: features.layer4.0 neuron: 482 value: 0.002 loss: -14.010 ATK Acc: 8.280 ATK Loss: 9.741 Norm: 15.997 Score: 10.861 Jaccard: 0.000 layer: features.layer3.1 neuron: 177 value: 0.123 loss: -344.448 ATK Acc: 9.980 ATK Loss: 7.316 Norm: 26.170 Score: 9.148 Jaccard: 0.000 layer: features.layer3.0 neuron: 59 value: 0.069 loss: -190.713 ATK Acc: 5.900 ATK Loss: 9.880 Norm: 16.001 Score: 11.000 Jaccard: 0.000 layer: features.layer3.0 neuron: 156 value: 0.006 loss: -200.982 ATK Acc: 8.300 ATK Loss: 9.708 Norm: 16.001 Score: 10.828 Jaccard: 0.000 layer: features.layer2.1 neuron: 76 value: 0.086 loss: -5684.504 ATK Acc: 11.910 ATK Loss: 6.007 Norm: 100.639 Score: 13.051 Jaccard: 0.000 Label: 8 loss: -344.448 ATK Acc: 9.980 ATK loss: 7.316 Norm: 26.170 Jaccard: 0.000 Score: 9.148 label: 9 layer: classifier.fc neuron: 9 value: 10.057 loss: -59.108 ATK Acc: 8.270 ATK Loss: 9.627 Norm: 16.003 Score: 10.748 Jaccard: 0.000 layer: features.layer4.1 neuron: 205 value: 0.001 loss: -187.362 ATK Acc: 8.000 ATK Loss: 9.737 Norm: 16.000 Score: 10.857 Jaccard: 0.000 layer: features.layer4.1 neuron: 65 value: 0.001 loss: -178.531 ATK Acc: 8.770 ATK Loss: 9.555 Norm: 15.998 Score: 10.674 Jaccard: 0.000 layer: features.layer4.0 neuron: 30 value: 0.008 loss: -75.536 ATK Acc: 7.240 ATK Loss: 9.707 Norm: 15.997 Score: 10.826 Jaccard: 0.133 layer: features.layer4.0 neuron: 319 value: 0.008 loss: -22.479 ATK Acc: 8.610 ATK Loss: 9.708 Norm: 15.997 Score: 10.827 Jaccard: 0.000 layer: features.layer4.0 neuron: 286 value: 0.006 loss: -21.125 ATK Acc: 8.550 ATK Loss: 9.683 Norm: 15.997 Score: 10.803 Jaccard: 0.000 layer: features.layer4.0 neuron: 8 value: 0.004 loss: -21.700 ATK Acc: 8.610 ATK Loss: 9.695 Norm: 15.995 Score: 10.815 Jaccard: 0.000 layer: features.layer3.1 neuron: 211 value: 0.089 loss: -162.960 ATK Acc: 5.450 ATK Loss: 9.559 Norm: 16.001 Score: 10.679 Jaccard: 0.000 layer: features.layer3.0 neuron: 186 value: 0.025 loss: -192.531 ATK Acc: 8.840 ATK Loss: 9.613 Norm: 16.001 Score: 10.733 Jaccard: 0.000 layer: features.layer3.0 neuron: 122 value: 0.007 loss: -180.473 ATK Acc: 7.670 ATK Loss: 9.894 Norm: 16.002 Score: 11.014 Jaccard: 0.000 layer: features.layer2.1 neuron: 62 value: 0.307 loss: -4121.620 ATK Acc: 2.020 ATK Loss: 6.662 Norm: 100.520 Score: 13.699 Jaccard: 0.000 layer: features.layer2.1 neuron: 113 value: 0.097 loss: -7281.228 ATK Acc: 92.000 ATK Loss: 0.263 Norm: 100.184 Score: 7.276 Jaccard: 0.000 layer: features.layer2.0 neuron: 113 value: 0.272 loss: -6832.499 ATK Acc: 95.960 ATK Loss: 0.172 Norm: 100.294 Score: 7.192 Jaccard: 0.000 Label: 9 loss: -6832.499 ATK Acc: 95.960 ATK loss: 0.172 Norm: 100.294 Jaccard: 0.000 Score: 7.192 Score: [7.156167772364617, 9.457871842956543, 4.543633675146103, 7.217127031528951, 5.883398698425293, 10.027698394012452, 10.763735179138184, 9.409244415664674, 9.147680532836914, 7.1922257972240455] Score MAD: tensor([0.0213, 0.7829, 0.9341, 0.0000, 0.4660, 0.9819, 1.2391, 0.7659, 0.6745, 0.0087])