LLM-Empowered State Representation for Reinforcement Learning

TL;DR

This paper introduces LESR, a novel method that leverages large language models to generate task-related state representations, significantly improving sample efficiency and performance in reinforcement learning tasks.

Contribution

LESR is the first approach to use LLMs for autonomous generation of task-specific state representations in reinforcement learning.

Findings

LESR outperforms baselines by 29% in Mujoco tasks.

LESR achieves 30% higher success rates in Gym-Robotics.

LESR demonstrates high sample efficiency in experiments.

Abstract

Conventional state representations in reinforcement learning often omit critical task-related details, presenting a significant challenge for value networks in establishing accurate mappings from states to task rewards. Traditional methods typically depend on extensive sample learning to enrich state representations with task-specific information, which leads to low sample efficiency and high time costs. Recently, surging knowledgeable large language models (LLM) have provided promising substitutes for prior injection with minimal human intervention. Motivated by this, we propose LLM-Empowered State Representation (LESR), a novel approach that utilizes LLM to autonomously generate task-related state representation codes which help to enhance the continuity of network mappings and facilitate efficient training. Experimental results demonstrate LESR exhibits high sample efficiency and outperforms state-of-the-art baselines by an average of 29% in accumulated reward in Mujoco tasks and 30% in success rates in Gym-Robotics tasks.