On Rollouts in Model-Based Reinforcement Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Frauenknecht, Bernd, Subhasish, Devdutt, Solowjow, Friedrich, Trimpe, Sebastian
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915233025490944
author Frauenknecht, Bernd
Subhasish, Devdutt
Solowjow, Friedrich
Trimpe, Sebastian
author_facet Frauenknecht, Bernd
Subhasish, Devdutt
Solowjow, Friedrich
Trimpe, Sebastian
contents Model-based reinforcement learning (MBRL) seeks to enhance data efficiency by learning a model of the environment and generating synthetic rollouts from it. However, accumulated model errors during these rollouts can distort the data distribution, negatively impacting policy learning and hindering long-term planning. Thus, the accumulation of model errors is a key bottleneck in current MBRL methods. We propose Infoprop, a model-based rollout mechanism that separates aleatoric from epistemic model uncertainty and reduces the influence of the latter on the data distribution. Further, Infoprop keeps track of accumulated model errors along a model rollout and provides termination criteria to limit data corruption. We demonstrate the capabilities of Infoprop in the Infoprop-Dyna algorithm, reporting state-of-the-art performance in Dyna-style MBRL on common MuJoCo benchmark tasks while substantially increasing rollout length and data quality.
format Preprint
id arxiv_https___arxiv_org_abs_2501_16918
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle On Rollouts in Model-Based Reinforcement Learning
Frauenknecht, Bernd
Subhasish, Devdutt
Solowjow, Friedrich
Trimpe, Sebastian
Machine Learning
Model-based reinforcement learning (MBRL) seeks to enhance data efficiency by learning a model of the environment and generating synthetic rollouts from it. However, accumulated model errors during these rollouts can distort the data distribution, negatively impacting policy learning and hindering long-term planning. Thus, the accumulation of model errors is a key bottleneck in current MBRL methods. We propose Infoprop, a model-based rollout mechanism that separates aleatoric from epistemic model uncertainty and reduces the influence of the latter on the data distribution. Further, Infoprop keeps track of accumulated model errors along a model rollout and provides termination criteria to limit data corruption. We demonstrate the capabilities of Infoprop in the Infoprop-Dyna algorithm, reporting state-of-the-art performance in Dyna-style MBRL on common MuJoCo benchmark tasks while substantially increasing rollout length and data quality.
title On Rollouts in Model-Based Reinforcement Learning
topic Machine Learning
url https://arxiv.org/abs/2501.16918