REvolve: Reward Evolution with Large Language Models using Human Feedback

TL;DR

REvolve leverages large language models guided by human feedback to automatically generate and refine reward functions for reinforcement learning in complex, subjective tasks, improving agent performance.

Contribution

This work introduces REvolve, an evolutionary framework that uses human feedback to guide LLMs in designing reward functions for challenging RL tasks.

Findings

Agents trained with REvolve rewards outperform baselines.

LLMs can encode implicit human knowledge into reward functions.

Human feedback effectively guides reward evolution.

Abstract

Designing effective reward functions is crucial to training reinforcement learning (RL) algorithms. However, this design is non-trivial, even for domain experts, due to the subjective nature of certain tasks that are hard to quantify explicitly. In recent works, large language models (LLMs) have been used for reward generation from natural language task descriptions, leveraging their extensive instruction tuning and commonsense understanding of human behavior. In this work, we hypothesize that LLMs, guided by human feedback, can be used to formulate reward functions that reflect human implicit knowledge. We study this in three challenging settings -- autonomous driving, humanoid locomotion, and dexterous manipulation -- wherein notions of ``good" behavior are tacit and hard to quantify. To this end, we introduce REvolve, a truly evolutionary framework that uses LLMs for reward design in RL. REvolve generates and refines reward functions by utilizing human feedback to guide the evolution process, effectively translating implicit human knowledge into explicit reward functions for training (deep) RL agents. Experimentally, we demonstrate that agents trained on REvolve-designed rewards outperform other state-of-the-art baselines.