If I understand correctly, the discount factor should be between 0-1 and is squared at each time step. Is the reward decay the same as the discount factor or is it the rate of decreasing the potential reward for each time step such as the square of the discount factor? When I look this question up only results for either discount factor or reward decay is mentioned but not both and never the difference between.
Is there a difference between a discount factor and reward decay in reinforcement learning?
$\begingroup$ I think these are the same concepts. Reward decay refers to the rewards being... decayed at each successive time step, that is decreasing in value, while the discount factor (often times $\gamma$) is the number determining by how much you decay the reward at each step (relatively). $\endgroup$– user2974951Jun 10, 2021 at 10:23
I think Reward decay decreases with each time step whereas the discount factor is a fixed number for the whole episode.
$\begingroup$ This does not provide an answer to the question. Once you have sufficient reputation you will be able to comment on any post; instead, provide answers that don't require clarification from the asker. - From Review $\endgroup$ Mar 16 at 8:29