$**$Part A$**$: Suppose each of the weights is initialized to $W$^$k $=1\mathrm{.}0$$ and each bias is initialized to $b$^$k $=-0\mathrm{.}5$$$.$ Use forward propagation to find the activities and activations associated with each hidden and output neuron for the training example $$($x$,$ y$)=(0\mathrm{.}5,0)$$$.$ Show your work. Answer the Peer Review question about this section. $**$Part B$**$: Use Back$-$Propagation to compute the weight and bias derivatives $$\backslash$partial $\backslash$ell $/\backslash$partial W$^$k$ and $$\backslash$partial $\backslash$ell $/\backslash$partial b$^$k$ for $k$=1,2,3$$$.$ Show all work. Answer the Peer Review question about this section. $**$PART C$**$ Implement following activation functions:

Formulas for activation functions

$*$ Relu: f$($$x$$)=$ max$(0,$ $x$$)$

$*$ Sigmoid: f$($$x$$)=$ $$\backslash$frac$\{1\}\{1+$ e$^\{-$x$\}\}$$

$*$ Softmax: f$($$x$\_$i$$)=$ $$\backslash$frac$\{$e$^$x$\_$i$\}\{\backslash$sum$\_\{$j$=1\}^\{$n$\}$ e$^\{$x$\_$j$\}\}$$