liveBook · Manning

Reward-to-go

Reward-to-go is the discounted cumulative reward from a specified decision point through the remainder of a trajectory.

How 2 Manning books explain it

Applied Reinforcement Learning

Return-to-go is the discounted sum of rewards from a specified decision time onward in a trajectory.

How it is used
Estimates policy gradients from sampled trajectories using rewards relevant to each action.
How deep it goes
Defines the discounted sum and explains its credit-assignment rationale, while leaving detailed calculations to a separate walkthrough.
What you can do after
Understand why gradient estimates use future rewards rather than the full trajectory return.

Read how Applied Reinforcement Learning explains reward-to-go →

Reinforcement Learning from Human Feedback

Reward-to-go is the cumulative future reward expected from a specified point in a trial or trajectory.

How it is used
Frames reinforcement-learning optimization around estimating cumulative future reward from a point in a trial.
How deep it goes
Defines cumulative future reward, introduces discounting, and connects estimation to Bellman recursion and temporal-difference updates.
Examples
Uses Q-learning, state–action value estimates, and Backgammon as concrete anchors for future-reward estimation.
What you can do after
Evaluate how RL methods estimate and update future reward under Markov decision process formulations.

Read how Reinforcement Learning from Human Feedback explains reward-to-go →

Related concepts

Questions about reward-to-go

What is reward-to-go?

Reward-to-go is the discounted cumulative reward from a specified decision point through the remainder of a trajectory.

Which Manning books explain reward-to-go?

Applied Reinforcement Learning and Reinforcement Learning from Human Feedback.

These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.