What implicit diferentiation does, is the same thing as the normal diferentiation. You diferentiate on both sides in respect to one variable, and the derivative of the other variable that depends on the first appears isolated in normal diferentiation, and not isolated in implicit, you only need to isolate it.
In your function, you first diferentiate on both sides in respect to x, you can of course isolate y, wich is easy in this function, but lets do it in the way the problem wants to.
Diferentiating on both sides, what have you got?