About 1x1 convolution

Is there any idea about how to deal with 1x1 convolution for channel alignment for parameter sharing CNN? Since the size changes as the graph change.

The issue names no file, test, or entry point to inspect. It asks for an approach to 1x1 convolution and channel alignment in a parameter-sharing CNN, so the first step would be to locate that implementation and establish the expected behavior; completion cannot be defined until an approach is agreed.

Written by the indexing model from the issue text.