Saved in:
Bibliographic Details
Main Authors: Georgiev, Ignat, Srinivasan, Krishnan, Xu, Jie, Heiden, Eric, Garg, Animesh
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2405.17784
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913376058212352
author Georgiev, Ignat
Srinivasan, Krishnan
Xu, Jie
Heiden, Eric
Garg, Animesh
author_facet Georgiev, Ignat
Srinivasan, Krishnan
Xu, Jie
Heiden, Eric
Garg, Animesh
contents Model-Free Reinforcement Learning (MFRL), leveraging the policy gradient theorem, has demonstrated considerable success in continuous control tasks. However, these approaches are plagued by high gradient variance due to zeroth-order gradient estimation, resulting in suboptimal policies. Conversely, First-Order Model-Based Reinforcement Learning (FO-MBRL) methods employing differentiable simulation provide gradients with reduced variance but are susceptible to sampling error in scenarios involving stiff dynamics, such as physical contact. This paper investigates the source of this error and introduces Adaptive Horizon Actor-Critic (AHAC), an FO-MBRL algorithm that reduces gradient error by adapting the model-based horizon to avoid stiff dynamics. Empirical findings reveal that AHAC outperforms MFRL baselines, attaining 40% more reward across a set of locomotion tasks and efficiently scaling to high-dimensional control environments with improved wall-clock-time efficiency.
format Preprint
id arxiv_https___arxiv_org_abs_2405_17784
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Adaptive Horizon Actor-Critic for Policy Learning in Contact-Rich Differentiable Simulation
Georgiev, Ignat
Srinivasan, Krishnan
Xu, Jie
Heiden, Eric
Garg, Animesh
Machine Learning
Artificial Intelligence
Model-Free Reinforcement Learning (MFRL), leveraging the policy gradient theorem, has demonstrated considerable success in continuous control tasks. However, these approaches are plagued by high gradient variance due to zeroth-order gradient estimation, resulting in suboptimal policies. Conversely, First-Order Model-Based Reinforcement Learning (FO-MBRL) methods employing differentiable simulation provide gradients with reduced variance but are susceptible to sampling error in scenarios involving stiff dynamics, such as physical contact. This paper investigates the source of this error and introduces Adaptive Horizon Actor-Critic (AHAC), an FO-MBRL algorithm that reduces gradient error by adapting the model-based horizon to avoid stiff dynamics. Empirical findings reveal that AHAC outperforms MFRL baselines, attaining 40% more reward across a set of locomotion tasks and efficiently scaling to high-dimensional control environments with improved wall-clock-time efficiency.
title Adaptive Horizon Actor-Critic for Policy Learning in Contact-Rich Differentiable Simulation
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2405.17784