Dr. Strategy: Model-Based Generalist Agents with Strategic Dreaming

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hamed, Hany, Kim, Subin, Kim, Dongyeong, Yoon, Jaesik, Ahn, Sungjin
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914822646398976
author Hamed, Hany
Kim, Subin
Kim, Dongyeong
Yoon, Jaesik
Ahn, Sungjin
author_facet Hamed, Hany
Kim, Subin
Kim, Dongyeong
Yoon, Jaesik
Ahn, Sungjin
contents Model-based reinforcement learning (MBRL) has been a primary approach to ameliorating the sample efficiency issue as well as to make a generalist agent. However, there has not been much effort toward enhancing the strategy of dreaming itself. Therefore, it is a question whether and how an agent can "dream better" in a more structured and strategic way. In this paper, inspired by the observation from cognitive science suggesting that humans use a spatial divide-and-conquer strategy in planning, we propose a new MBRL agent, called Dr. Strategy, which is equipped with a novel Dreaming Strategy. The proposed agent realizes a version of divide-and-conquer-like strategy in dreaming. This is achieved by learning a set of latent landmarks and then utilizing these to learn a landmark-conditioned highway policy. With the highway policy, the agent can first learn in the dream to move to a landmark, and from there it tackles the exploration and achievement task in a more focused way. In experiments, we show that the proposed model outperforms prior pixel-based MBRL methods in various visually complex and partially observable navigation tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2402_18866
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Dr. Strategy: Model-Based Generalist Agents with Strategic Dreaming
Hamed, Hany
Kim, Subin
Kim, Dongyeong
Yoon, Jaesik
Ahn, Sungjin
Machine Learning
Model-based reinforcement learning (MBRL) has been a primary approach to ameliorating the sample efficiency issue as well as to make a generalist agent. However, there has not been much effort toward enhancing the strategy of dreaming itself. Therefore, it is a question whether and how an agent can "dream better" in a more structured and strategic way. In this paper, inspired by the observation from cognitive science suggesting that humans use a spatial divide-and-conquer strategy in planning, we propose a new MBRL agent, called Dr. Strategy, which is equipped with a novel Dreaming Strategy. The proposed agent realizes a version of divide-and-conquer-like strategy in dreaming. This is achieved by learning a set of latent landmarks and then utilizing these to learn a landmark-conditioned highway policy. With the highway policy, the agent can first learn in the dream to move to a landmark, and from there it tackles the exploration and achievement task in a more focused way. In experiments, we show that the proposed model outperforms prior pixel-based MBRL methods in various visually complex and partially observable navigation tasks.
title Dr. Strategy: Model-Based Generalist Agents with Strategic Dreaming
topic Machine Learning
url https://arxiv.org/abs/2402.18866