How to extract learned weights by channel

Follow this issue,

Is there a way to get the sum of by channel dim? Thanks!

I assume you mean the in_channel dimension of a conv layer by “channel dim”?
If so, this would work:

conv = nn.Conv2d(3, 6, 3, 1, 1)
conv.weight.sum([0, 2, 3])
> tensor([-0.2552,  0.2478,  0.3706], grad_fn=<SumBackward1>)

Thanks for your reply. Actually, I would like to explore the importance of the channel-wise features by summing up the channel-wise weights for a group=3 Conv.

I tried to calculate a group=3 3x3 conv’s sum of weights by this,

            mid1_1 = net.conv1[-3][1]//3
            mid1_2 = net.conv1[-3][1]//3*2
            conv1_g1_ws = torch.sum(net.conv1[-3][:mid1_1, :mid1_1, :, :])
            conv1_g2_ws = torch.sum(net.conv1[-3][mid1_1:mid1_2, mid1_1:mid1_2, :, :])
            conv1_g3_ws = torch.sum(net.conv1[-3][mid1_2:, mid1_2:, :, :])

Am I right? Many thanks for your advice.