Generating and Evolving Reward Functions for Highway Driving with Large Language Models

TL;DR

This paper presents a novel framework that leverages Large Language Models to automatically generate and refine reward functions for autonomous highway driving, significantly improving success rates over handcrafted rewards.

Contribution

Introducing a framework that uses LLMs to generate and evolve reward functions for autonomous driving, reducing manual effort and enhancing performance.

Findings

Achieved 22% higher success rate than handcrafted reward functions.

Demonstrated effective reward function generation in highway scenarios.

Improved development productivity for autonomous driving systems.

Abstract

Reinforcement Learning (RL) plays a crucial role in advancing autonomous driving technologies by maximizing reward functions to achieve the optimal policy. However, crafting these reward functions has been a complex, manual process in many practices. To reduce this complexity, we introduce a novel framework that integrates Large Language Models (LLMs) with RL to improve reward function design in autonomous driving. This framework utilizes the coding capabilities of LLMs, proven in other areas, to generate and evolve reward functions for highway scenarios. The framework starts with instructing LLMs to create an initial reward function code based on the driving environment and task descriptions. This code is then refined through iterative cycles involving RL training and LLMs' reflection, which benefits from their ability to review and improve the output. We have also developed a specific prompt template to improve LLMs' understanding of complex driving simulations, ensuring the generation of effective and error-free code. Our experiments in a highway driving simulator across three traffic configurations show that our method surpasses expert handcrafted reward functions, achieving a 22% higher average success rate. This not only indicates safer driving but also suggests significant gains in development productivity.