Definition. Return [sutton2018reinforcement, Section 3.3] [ftip-0030]

For an episode ending at time \(N\), the return from decision time \(t\) is

\[ G_t=\sum _{k=t}^{N-1}\gamma ^{k-t}R_{k+1}. \]

When all intermediate rewards vanish, every return is determined by the terminal reward, up to discounting. Return is a random variable under the policy and environment law; its expectation defines value.