Learning Action Translator for Meta Reinforcement Learning on Sparse-Reward Tasks

TL;DR

This paper introduces an action translator for meta reinforcement learning that enhances sample efficiency and performance on sparse-reward, long-horizon tasks by enabling better policy transfer and exploration.

Contribution

It proposes a novel objective for learning an action translator, with theoretical guarantees, integrated with meta-RL algorithms to improve learning efficiency on challenging tasks.

Findings

Improved sample efficiency on sparse-reward tasks.

Enhanced policy transfer via the action translator.

Empirical performance gains over baseline meta-RL methods.

Abstract

Meta reinforcement learning (meta-RL) aims to learn a policy solving a set of training tasks simultaneously and quickly adapting to new tasks. It requires massive amounts of data drawn from training tasks to infer the common structure shared among tasks. Without heavy reward engineering, the sparse rewards in long-horizon tasks exacerbate the problem of sample efficiency in meta-RL. Another challenge in meta-RL is the discrepancy of difficulty level among tasks, which might cause one easy task dominating learning of the shared policy and thus preclude policy adaptation to new tasks. This work introduces a novel objective function to learn an action translator among training tasks. We theoretically verify that the value of the transferred policy with the action translator can be close to the value of the source policy and our objective function (approximately) upper bounds the value difference. We propose to combine the action translator with context-based meta-RL algorithms for better data collection and more efficient exploration during meta-training. Our approach empirically improves the sample efficiency and performance of meta-RL algorithms on sparse-reward tasks.