arXiv cs.AI· Michal P. Podolinsky, Yunhao Yang, Daniel Little, Hong Lu, Rohan Siva, Olin Goudey, Neel P. Bhatt, Matthias Scheutz, Ufuk Topcu·· 12 小时前AI 评分34
VLA-Tug：通过校准潜在探针实现视觉-语言-动作模型的失败预测与分类
Failure Prediction and Classification in Vision-Language-Action Models Via Calibrated Latent Probing
AI 导读
VLA-Tug 是一种运行在 VLA 主干潜在层上的轻量探针，可在最小推理开销下实现早期失败检测，并约束漏检率与误报率。在相同误报率下，其早期失败检测能力比最强学习基线提升最多 30%；迁移到未见任务时，漏检与误报边界在最多 86% 的运行中仍然有效，而基线最多为 21%。
正文 · 原文
Abstract:Vision-Language-Action models (VLAs) enable general-purpose embodied intelligence, but their performance is brittle under distribution shift, where unanticipated failures can damage equipment or endanger people. Deploying VLAs in real-world systems therefore requires a runtime monitor that is timely, calibrated, and informative while being efficient enough to run on the edge. Existing methods fail to fulfill these requirements. We present VLA-Tug, a lightweight probe over the latent layers of a VLA backbone that improves early failure detection and bounds the missed-detection and false-alarm rates, with minimal inference overhead. We show that supervising on the pre-failure segment of a trajectory increases early failure detection at matched false-positive rates by up to 30% over the strongest learned baseline. Under shift to unseen tasks, the bounds on missed-detection and false-alarm rates remain valid in up to 86% of runs, versus at most 21% for baselines. Unlike existing methods, VLA-Tug predicts the failure mode to enable corrective actions, raising top-3 accuracy on unseen tasks by up to 44% over a majority baseline. VLA-Tug enables real-time monitoring where the well-timed and informative probe supports lightweight steering that improves task success. We release a dataset of 3,180 annotated manipulation episodes to support future work on latent failure detection in VLAs at this https URL.
| Comments: | Code, videos, and dataset available at this https URL |
| Subjects: | Robotics (cs.RO); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.08425 [cs.RO] |
| (or arXiv:2611.08425v1 [cs.RO] for this version) | |
| https://doi.org/10.48550/arXiv.2611.08425 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Michal Podolinsky [view email]
[v1]
Thu, 5 Nov 2026 18:21:07 UTC (20,516 KB)
来源：arXiv cs.AI · arxiv.org