Abstract:The ability to efficiently and reliably learn new tasks has been a foundational challenge in robotics. Vision-Language-Action (VLA) models have demonstrated strong generalization across diverse manipulation tasks, yet pretrained policies consistently fall short of the reliability required for real-world deployment. Reinforcement learning (RL) fine-tuning offers a promising path to bridge this gap, but existing approaches either train from scratch without fully leveraging pretrained priors, or fine-tune VLAs without achieving the sample efficiency and success rates that practical deployment demands. We present EXPO-FT, a system for stable, sample-efficient RL finetuning of pretrained VLA policies that closes this gap. Our system solves a suite of challenging manipulation tasks, including routing string lights and inserting the plug to light it up, striking a pool ball into a pocket, and inserting a flower into a wine bottle, each requiring combinations of high precision, dynamic actions, and robustness to varied initial states. Our system achieves perfect task performance (30/30 successes) across all evaluated tasks within an average of 19.1 minutes of online robot data, outperforming both prior RL-from-scratch and VLA finetuning approaches. We release an open-source codebase with the aim of facilitating broader adoption of RL finetuning of VLA models in robotics.
| Subjects: | Robotics (cs.RO); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2605.25477 [cs.RO] |
| (or arXiv:2605.25477v1 [cs.RO] for this version) | |
| https://doi.org/10.48550/arXiv.2605.25477 arXiv-issued DOI via DataCite (pending registration) |
From: Perry Dong [view email]
[v1]
Mon, 25 May 2026 06:31:03 UTC (27,059 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。