Teachable Reinforcement Learning via Advice Distillation
This addresses the problem of reducing human supervision in training automated agents for complex tasks, offering a novel approach that is not incremental but introduces a new supervision paradigm.
The paper tackles the challenge of training agents in interactive environments by introducing a 'teachable' paradigm where agents learn from structured advice from a teacher, resulting in agents acquiring new skills with significantly less human supervision than standard reinforcement learning and often less than imitation learning in puzzle-solving, navigation, and locomotion domains.
Training automated agents to complete complex tasks in interactive environments is challenging: reinforcement learning requires careful hand-engineering of reward functions, imitation learning requires specialized infrastructure and access to a human expert, and learning from intermediate forms of supervision (like binary preferences) is time-consuming and extracts little information from each human intervention. Can we overcome these challenges by building agents that learn from rich, interactive feedback instead? We propose a new supervision paradigm for interactive learning based on "teachable" decision-making systems that learn from structured advice provided by an external teacher. We begin by formalizing a class of human-in-the-loop decision making problems in which multiple forms of teacher-provided advice are available to a learner. We then describe a simple learning algorithm for these problems that first learns to interpret advice, then learns from advice to complete tasks even in the absence of human supervision. In puzzle-solving, navigation, and locomotion domains, we show that agents that learn from advice can acquire new skills with significantly less human supervision than standard reinforcement learning algorithms and often less than imitation learning.