excitationbp
excitationbp copied to clipboard
Visualizing how deep networks make decisions
I am very interested in this work. I am trying to do some new work on excitationbp, and there is a operation named "eb.use_eb()" to modify "torch.nn.functional". The functiona eb.use_eb()...
>>> import excitationbp as eb Traceback (most recent call last): File "", line 1, in File "/work/qinfeng/excitationbp/excitationbp/__init__.py", line 9, in from .functions.eb_convNd import * File "/work/qinfeng/excitationbp/excitationbp/functions/eb_convNd.py", line 322, in _bind_functions()...
Hi @greydanus I'm not sure whether it is caused by the maxpooling in my network. After feature extraction by CNN (B,C,14,14), My final layer : nn.MaxPool2d(14,1,0) -> view(output.size(0), self.num_classes) where...
Going to look into this in detail now, but got the following error while running the imagenet ipynb ``` RuntimeErrorTraceback (most recent call last) in () 8 prob_outputs_dog = Variable(torch.zeros(1,1000))...