2D convolution forward pass in NumPy, with parameter counts

0
Remix
  1. Given the input tensor shape, input channels, output channels, kernel size, stride and padding, how many parameters does a Conv2D layer have, with and without bias? What is the output shape?
  2. Implement the forward pass of a 2D convolution in NumPy, first with plain loops.
  3. Vectorize it (for example with an im2col-style reshaping) and check it against your loop version.
TeslaTechnical phone screenLast asked Dec 2025
No answer yet. Write the first one; the best answer gets endorsed.