RF-Agent: Automated Reward Function Design via Language Agent Tree Search

TL;DR

RF-Agent introduces a novel framework that uses language agent tree search with LLMs and MCTS to efficiently design reward functions for low-level control tasks, outperforming previous methods.

Contribution

The paper presents RF-Agent, a new approach that frames reward function design as a sequential decision process using LLMs and MCTS, improving search efficiency and utilization of feedback.

Findings

Outperforms existing reward design methods in 17 control tasks.

Enhances reward function optimization through better contextual reasoning.

Demonstrates significant improvements in complex control scenarios.

Abstract

Designing efficient reward functions for low-level control tasks is a challenging problem. Recent research aims to reduce reliance on expert experience by using Large Language Models (LLMs) with task information to generate dense reward functions. These methods typically rely on training results as feedback, iteratively generating new reward functions with greedy or evolutionary algorithms. However, they suffer from poor utilization of historical feedback and inefficient search, resulting in limited improvements in complex control tasks. To address this challenge, we propose RF-Agent, a framework that treats LLMs as language agents and frames reward function design as a sequential decision-making process, enhancing optimization through better contextual reasoning. RF-Agent integrates Monte Carlo Tree Search (MCTS) to manage the reward design and optimization process, leveraging the multi-stage contextual reasoning ability of LLMs. This approach better utilizes historical information and improves search efficiency to identify promising reward functions. Outstanding experimental results in 17 diverse low-level control tasks demonstrate the effectiveness of our method. The source code is available at https://github.com/deng-ai-lab/RF-Agent.