Decoupled Learning of Environment Characteristics for Safe Exploration

TL;DR

This paper introduces a method for reinforcement learning agents to decouple environment characteristics from task-specific features, enhancing transferability and safety during learning of new tasks.

Contribution

The paper proposes a novel decoupled learning approach that separates environment traits from task details, improving safe exploration and transferability in reinforcement learning.

Findings

Decoupled learning improves safety during task learning.

Agents better transfer skills across tasks within the same environment.

Enhanced ability to reuse prior knowledge safely.

Abstract

Reinforcement learning is a proven technique for an agent to learn a task. However, when learning a task using reinforcement learning, the agent cannot distinguish the characteristics of the environment from those of the task. This makes it harder to transfer skills between tasks in the same environment. Furthermore, this does not reduce risk when training for a new task. In this paper, we introduce an approach to decouple the environment characteristics from the task-specific ones, allowing an agent to develop a sense of survival. We evaluate our approach in an environment where an agent must learn a sequence of collection tasks, and show that decoupled learning allows for a safer utilization of prior knowledge.