$)(4$ points$)$ To update the weights based on the gradients we computed, we will

apply the gradient descent rule:

W$($t$+1)=$ W$($t$)\backslash$gamma

$$L

$$W

Using a learning rate $(\backslash$gamma $)$ of $0\mathrm{.}01,$ compute the weights of W$1$ after one update.

W

$(1)$

$1=$

w$11$ w$12$

w$13$ w$14$

$=$