Intelligent Switching for Reset-Free RL
This addresses a key limitation for deploying RL in real-world applications where resets are unavailable, though it is incremental as it builds on prior reset-free RL methods.
The paper tackles the problem of training reinforcement learning agents without manual resets by introducing an algorithm that intelligently switches between forward and backward agents based on confidence, achieving state-of-the-art performance on challenging environments.
In the real world, the strong episode resetting mechanisms that are needed to train agents in simulation are unavailable. The \textit{resetting} assumption limits the potential of reinforcement learning in the real world, as providing resets to an agent usually requires the creation of additional handcrafted mechanisms or human interventions. Recent work aims to train agents (\textit{forward}) with learned resets by constructing a second (\textit{backward}) agent that returns the forward agent to the initial state. We find that the termination and timing of the transitions between these two agents are crucial for algorithm success. With this in mind, we create a new algorithm, Reset Free RL with Intelligently Switching Controller (RISC) which intelligently switches between the two agents based on the agent's confidence in achieving its current goal. Our new method achieves state-of-the-art performance on several challenging environments for reset-free RL.