In reinforcement learning, how do we know the contribution $($i$.$e$.,$ the credit$)$ of an action to the final reward?

Group of answer choices

we can check the reward at that moment when the action is taken.

we can sum up all rewards in the future since that action

we can sum up all rewards up to that action

we can sum up all rewards in the future since that action, with each weighted by an exponentially decreasing factor.