Let W$1,$ b$1,$ W$2,$ b$2$ be trainable parameters. Consider the following operations of a neural network on an input $x,$ taken sequentially.

$y=W1x+b1$

$z=W2y+b2$

$p=$ Sigmoid $(z)$

$c=$ SoftMax $($p$)$

This implements a

Let W$1,$ b$1,$ W$2,$ b$2$ be trainable parameters. Consider the following operations of a neural network on an input $x,$ taken sequentially.

$y=W1x+b1$

$z=W2y+b2$

$p=$ Sigmoid $(z)$

$c=$ SoftMax $($p$)$

This implements a