EfficientZero V2: Mastering Discrete and Continuous Control with Limited Data

Sample efficiency remains a crucial challenge in applying Reinforcement Learning (RL) to real-world tasks. While recent algorithms have made significant strides in improving sample efficiency, none have achieved consistently superior performance across diverse domains. In this paper, we introduce Ef...

Full description

Saved in:
Bibliographic Details
Published inarXiv.org
Main Authors Wang, Shengjie, Liu, Shaohuai, Ye, Weirui, You, Jiacheng, Gao, Yang
Format Paper
LanguageEnglish
Published Ithaca Cornell University Library, arXiv.org 12.09.2024
Subjects
Algorithms
Online AccessGet full text

Cover

More Information
Summary:Sample efficiency remains a crucial challenge in applying Reinforcement Learning (RL) to real-world tasks. While recent algorithms have made significant strides in improving sample efficiency, none have achieved consistently superior performance across diverse domains. In this paper, we introduce EfficientZero V2, a general framework designed for sample-efficient RL algorithms. We have expanded the performance of EfficientZero to multiple domains, encompassing both continuous and discrete actions, as well as visual and low-dimensional inputs. With a series of improvements we propose, EfficientZero V2 outperforms the current state-of-the-art (SOTA) by a significant margin in diverse tasks under the limited data setting. EfficientZero V2 exhibits a notable advancement over the prevailing general algorithm, DreamerV3, achieving superior outcomes in 50 of 66 evaluated tasks across diverse benchmarks, such as Atari 100k, Proprio Control, and Vision Control.
ISSN:2331-8422