Is the output of a strided convolution layer smaller or larger than a regular convolution layer? Group of answer choices: $1)$ The output of a strided convolution layer can be either smaller or larger than the output of a regular convolution layer depending on the kernel size. $2)$ The output of a strided convolution layer is smaller than the output of a regular convolution layer because it reduces the spatial dimensions of the input by skipping input values. $3)$ The output of a strided convolution layer is larger than the output of a regular convolution layer because it skips more input values. $4)$ The output of a strided convolution layer is the same size as the output of a regular convolution layer because the stride only affects the number of computations, not the output size.