Towards Deploying VLA without Fine-Tuning: Plug-and-Play Inference-Time VLA Policy Steering via Embodied Evolutionary Diffusion

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Zhuo, Liu, Junjia, Dong, Zhipeng, Teng, Tao, Rouxel, Quentin, Caldwell, Darwin, Chen, Fei
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917412535795712
author Li, Zhuo
Liu, Junjia
Dong, Zhipeng
Teng, Tao
Rouxel, Quentin
Caldwell, Darwin
Chen, Fei
author_facet Li, Zhuo
Liu, Junjia
Dong, Zhipeng
Teng, Tao
Rouxel, Quentin
Caldwell, Darwin
Chen, Fei
contents Vision-Language-Action (VLA) models have demonstrated significant potential in real-world robotic manipulation. However, pre-trained VLA policies still suffer from substantial performance degradation during downstream deployment. Although fine-tuning can mitigate this issue, its reliance on costly demonstration collection and intensive computation makes it impractical in real-world settings. In this work, we introduce VLA-Pilot, a plug-and-play inference-time policy steering method for zero-shot deployment of pre-trained VLA without any additional fine-tuning or data collection. We evaluate VLA-Pilot on six real-world downstream manipulation tasks across two distinct robotic embodiments, encompassing both in-distribution and out-of-distribution scenarios. Experimental results demonstrate that VLA-Pilot substantially boosts the success rates of off-the-shelf pre-trained VLA policies, enabling robust zero-shot generalization to diverse tasks and embodiments. Experimental videos and code are available at: https://rip4kobe.github.io/vla-pilot/.
format Preprint
id arxiv_https___arxiv_org_abs_2511_14178
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Towards Deploying VLA without Fine-Tuning: Plug-and-Play Inference-Time VLA Policy Steering via Embodied Evolutionary Diffusion
Li, Zhuo
Liu, Junjia
Dong, Zhipeng
Teng, Tao
Rouxel, Quentin
Caldwell, Darwin
Chen, Fei
Robotics
Artificial Intelligence
Vision-Language-Action (VLA) models have demonstrated significant potential in real-world robotic manipulation. However, pre-trained VLA policies still suffer from substantial performance degradation during downstream deployment. Although fine-tuning can mitigate this issue, its reliance on costly demonstration collection and intensive computation makes it impractical in real-world settings. In this work, we introduce VLA-Pilot, a plug-and-play inference-time policy steering method for zero-shot deployment of pre-trained VLA without any additional fine-tuning or data collection. We evaluate VLA-Pilot on six real-world downstream manipulation tasks across two distinct robotic embodiments, encompassing both in-distribution and out-of-distribution scenarios. Experimental results demonstrate that VLA-Pilot substantially boosts the success rates of off-the-shelf pre-trained VLA policies, enabling robust zero-shot generalization to diverse tasks and embodiments. Experimental videos and code are available at: https://rip4kobe.github.io/vla-pilot/.
title Towards Deploying VLA without Fine-Tuning: Plug-and-Play Inference-Time VLA Policy Steering via Embodied Evolutionary Diffusion
topic Robotics
Artificial Intelligence
url https://arxiv.org/abs/2511.14178