LLM4RL: Enhancing Reinforcement Learning with Large Language Models

被引：0

作者：

Zhou, Jiehan ^{[1
,2
]}

Zhao, Yang ^{[1
]}

Liu, Jiahong ^{[1
]}

Dong, Peijun ^{[1
]}

Luo, Xiaoyu ^{[3
]}

Tao, Hang ^{[1
]}

Chang, Shi ^{[4
]}

Luo, Hanjiang ^{[1
]}

机构：

[1] Shandong Univ Sci & Technol, Qingdao, Peoples R China

[2] Univ Oulu, Fac Informat Technol & Elect Engn, Oulu, Finland

[3] Kyung Hee Univ, Seoul, South Korea

[4] Univ Western Ontario, Dept Comp Sci, London, ON, Canada

来源：

2024 IEEE CANADIAN CONFERENCE ON ELECTRICAL AND COMPUTER ENGINEERING, CCECE 2024 | 2024年

关键词：

Large Language Model; Autonomous Driving; Reinforcement Learning; LLMs-enhanced Applications;

D O I：

10.1109/CCECE59415.2024.10667224

中图分类号：

TP39 [计算机的应用];

学科分类号：

081203 ; 0835 ;

摘要：

Integrating large language models (LLMs) into reinforcement learning (RL) promises to enhance the learning performance. Traditional RL faces challenges in industrial settings, including complex environments, safety concerns, and multimodal data. As powerful tools for contextual learning and reasoning, LLMs can address issues inherent in traditional RL. This paper introduces a generic LLM4RL framework, and investigates how LLM4RL can improve learning performance in autonomous driving.

引用

页码：86 / 87

页数：2

共 8 条

[1] Cao YJ, 2024, Arxiv, DOI arXiv:2404.00282
[2] Chakraborty S, 2023, Arxiv, DOI arXiv:2303.07622
[3] Du Y., 2023, Guiding pretraining in reinforcement learning with large language models
[4] Exploring Bus Stop Mobility Pattern: A Multi-Pattern Deep Learning Prediction Framework
Kong, Xiangjie
Shen, Zhehui
Wang, Kailai
Shen, Guojiang
Fu, Yanjie
[J]. IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, 2024, 25 (07) : 6604 - 6616
[5] Kwon M, 2023, Arxiv, DOI arXiv:2303.00001
[6] Li H, 2024, Arxiv, DOI arXiv:2312.09238
[7] Lin J, 2024, Arxiv, DOI arXiv:2308.01399
[8] Pang JC, 2023, Arxiv, DOI arXiv:2302.09368

← 1 →