Saved in:
Bibliographic Details
Main Authors: Gupta, Shashank, Lee, Pilhwa
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2505.16741
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918491259404288
author Gupta, Shashank
Lee, Pilhwa
author_facet Gupta, Shashank
Lee, Pilhwa
contents Minimum attention applies the least action principle to changes of control concerning state and time, first proposed by Brockett. The involved regularization is highly relevant in emulating biological control, such as motor learning. We apply minimum attention in reinforcement learning (RL) as part of the rewards and investigate its connection to meta-learning and stabilization. Specifically, model-based meta-learning with minimum attention is explored in high-dimensional nonlinear dynamics. Ensemble-based model learning and gradient-based meta-policy learning are alternately performed. Empirically, the minimum attention does show outperforming competence in comparison to the state-of-the-art algorithms of model-free and model-based RL, i.e., fast adaptation in few shots and variance reduction from the perturbations of the model and environment. Furthermore, the minimum attention demonstrates an improvement in energy efficiency.
format Preprint
id arxiv_https___arxiv_org_abs_2505_16741
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Meta-reinforcement learning with minimum attention
Gupta, Shashank
Lee, Pilhwa
Machine Learning
Optimization and Control
Minimum attention applies the least action principle to changes of control concerning state and time, first proposed by Brockett. The involved regularization is highly relevant in emulating biological control, such as motor learning. We apply minimum attention in reinforcement learning (RL) as part of the rewards and investigate its connection to meta-learning and stabilization. Specifically, model-based meta-learning with minimum attention is explored in high-dimensional nonlinear dynamics. Ensemble-based model learning and gradient-based meta-policy learning are alternately performed. Empirically, the minimum attention does show outperforming competence in comparison to the state-of-the-art algorithms of model-free and model-based RL, i.e., fast adaptation in few shots and variance reduction from the perturbations of the model and environment. Furthermore, the minimum attention demonstrates an improvement in energy efficiency.
title Meta-reinforcement learning with minimum attention
topic Machine Learning
Optimization and Control
url https://arxiv.org/abs/2505.16741