The Evolving Landscape of LLM- and VLM-Integrated Reinforcement Learning

TL;DR

This survey reviews how Large Language Models and Vision-Language Models are integrated into reinforcement learning to address challenges like knowledge gaps and planning, proposing a taxonomy and future research directions.

Contribution

It introduces a taxonomy categorizing LLM/VLM-assisted RL approaches into agent, planner, and reward roles, and consolidates current research with future challenges.

Findings

Categorizes LLM/VLM roles in RL as agent, planner, reward

Identifies open problems like grounding and bias mitigation

Provides a framework for future integration of multimodal models in RL

Abstract

Reinforcement learning (RL) has shown impressive results in sequential decision-making tasks. Meanwhile, Large Language Models (LLMs) and Vision-Language Models (VLMs) have emerged, exhibiting impressive capabilities in multimodal understanding and reasoning. These advances have led to a surge of research integrating LLMs and VLMs into RL. In this survey, we review representative works in which LLMs and VLMs are used to overcome key challenges in RL, such as lack of prior knowledge, long-horizon planning, and reward design. We present a taxonomy that categorizes these LLM/VLM-assisted RL approaches into three roles: agent, planner, and reward. We conclude by exploring open problems, including grounding, bias mitigation, improved representations, and action advice. By consolidating existing research and identifying future directions, this survey establishes a framework for integrating LLMs and VLMs into RL, advancing approaches that unify natural language and visual understanding with sequential decision-making.