RobustVLA: Robustness-Aware Reinforcement Post-Training for Vision-Language-Action Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhang, Hongyin, Zhang, Shuo, Jin, Junxi, Zeng, Qixin, Li, Runze, Wang, Donglin
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909936551723008
author Zhang, Hongyin
Zhang, Shuo
Jin, Junxi
Zeng, Qixin
Li, Runze
Wang, Donglin
author_facet Zhang, Hongyin
Zhang, Shuo
Jin, Junxi
Zeng, Qixin
Li, Runze
Wang, Donglin
contents Vision-Language-Action (VLA) models have recently emerged as powerful general-purpose policies for robotic manipulation, benefiting from large-scale multi-modal pre-training. However, they often fail to generalize reliably in out-of-distribution deployments, where unavoidable disturbances such as observation noise, sensor errors, or actuation perturbations become prevalent. While recent Reinforcement Learning (RL)-based post-training provides a practical means to adapt pre-trained VLA models, existing methods mainly emphasize reward maximization and overlook robustness to environmental uncertainty. In this work, we introduce RobustVLA, a lightweight online RL post-training method designed to explicitly enhance the resilience of VLA models. Through a systematic robustness analysis, we identify two key regularizations: Jacobian regularization, which mitigates sensitivity to observation noise, and smoothness regularization, which stabilizes policies under action perturbations. Extensive experiments across diverse robotic environments demonstrate that RobustVLA significantly outperforms prior state-of-the-art methods in robustness and reliability. Our results highlight the importance of principled robustness-aware RL post-training as a key step toward improving the reliability and robustness of VLA models.
format Preprint
id arxiv_https___arxiv_org_abs_2511_01331
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle RobustVLA: Robustness-Aware Reinforcement Post-Training for Vision-Language-Action Models
Zhang, Hongyin
Zhang, Shuo
Jin, Junxi
Zeng, Qixin
Li, Runze
Wang, Donglin
Robotics
Machine Learning
Vision-Language-Action (VLA) models have recently emerged as powerful general-purpose policies for robotic manipulation, benefiting from large-scale multi-modal pre-training. However, they often fail to generalize reliably in out-of-distribution deployments, where unavoidable disturbances such as observation noise, sensor errors, or actuation perturbations become prevalent. While recent Reinforcement Learning (RL)-based post-training provides a practical means to adapt pre-trained VLA models, existing methods mainly emphasize reward maximization and overlook robustness to environmental uncertainty. In this work, we introduce RobustVLA, a lightweight online RL post-training method designed to explicitly enhance the resilience of VLA models. Through a systematic robustness analysis, we identify two key regularizations: Jacobian regularization, which mitigates sensitivity to observation noise, and smoothness regularization, which stabilizes policies under action perturbations. Extensive experiments across diverse robotic environments demonstrate that RobustVLA significantly outperforms prior state-of-the-art methods in robustness and reliability. Our results highlight the importance of principled robustness-aware RL post-training as a key step toward improving the reliability and robustness of VLA models.
title RobustVLA: Robustness-Aware Reinforcement Post-Training for Vision-Language-Action Models
topic Robotics
Machine Learning
url https://arxiv.org/abs/2511.01331