There is a variant of Standard Conjecture D for projective varieties over finite fields. It claims that rational and homological equivalences are equivalent on cycles after tensoring with $\mathbb{Q}$. This is often referred to as Beilinson's conjecture. Is it known whether this is true for product of curves?

Another related question: Is it known whether this conjecture is birationally invariant?