Intentional Updates for Streaming Reinforcement Learning
This work addresses instability in streaming RL (batch size=1) for practitioners needing sample-efficient, online learning without replay buffers.
The authors propose 'intentional updates' for streaming reinforcement learning, where the step size is chosen to achieve a specified outcome (e.g., fixed fractional reduction of TD error or bounded policy change). Their methods achieve state-of-the-art streaming performance, often matching batch and replay-buffer approaches.
In gradient-based learning, a step size chosen in parameter units does not produce a predictable per-step change in function output. This often leads to instability in the streaming setting (i.e., batch size=1), where stochasticity is not averaged out and update magnitudes can momentarily become arbitrarily big or small. Instead, we propose intentional updates: first specify the intended outcome of an update and then solve for the step size that approximately achieves it. This strategy has precedent in online supervised linear regression via Normalized Least Mean Squares algorithm, which selects a step size to yield a specified change in the function output proportional to the current error. We extend this principle to streaming deep reinforcement learning by defining appropriate intended outcomes: Intentional TD aims for a fixed fractional reduction of the TD error, and Intentional Policy Gradient aims for a bounded per-step change in the policy, limiting local KL divergence. We propose practical algorithms combining eligibility traces and diagonal scaling. Empirically, these methods yield state-of-the-art streaming performance, frequently performing on par with batch and replay-buffer approaches.