الملخص
Vision-language-action models (VLAs) are strikingly sensitive to instruction phrasing and do not inherit the language robustness of the vision-language models they are built on. A one-word edit can move success by tens of points: $π_{0.5}$ turns on a LIBERO stove 100% of the time for "switch on the stove" and 2% for "switch on the hot plate", and a $π_0$ checkpoint finetuned with rephrase augmentation still shows swings of up to 61 points. We characterize this sensitivity with statistically tested single-edit swings and an oracle phrase search, which shows that phrasing alone nearly closes the 21-point gap between in-distribution and out-of-distribution tasks. We then reduce it without modifying the policy. Because the sensitivity is systematic, it can be expressed as explicit rules: we score many phrasings of a few training tasks, have a large language model distill the evidence into ten to twenty rephrasing rules, and at deployment rewrite each incoming instruction once under these rules. The rules improve the frozen $π_0$ by 16 to 27% relative on twelve held-out tasks across adversarial, VLM-generated, and human-generated phrasings, with gains concentrated on out-of-distribution tasks. The pipeline replicates on $π_{0.5}$ and LIBERO, lifting in-finetune success from 93.6% to 97.8%. The method requires no retraining and no per-step verification, and applies zero-shot to unseen tasks and instructions. Project website: https://sttawm.github.io/rephrase-before-you-act
الكلمات المفتاحية
الموضوع
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Watts, M., & Cui, Y. (2026). Rephrase Before You Act: Characterizing and Mitigating Language Sensitivity in Vision-Language-Action Models. https://omanscience.com/ar/articles/rephrase-before-you-act-characterizing-and-mitigating-language-sensitivity-in-vision-language-action-models
MLA 9
Watts, Mikey, and Yuchen Cui. "Rephrase Before You Act: Characterizing and Mitigating Language Sensitivity in Vision-Language-Action Models." https://omanscience.com/ar/articles/rephrase-before-you-act-characterizing-and-mitigating-language-sensitivity-in-vision-language-action-models.
شيكاغو (المؤلف–التاريخ)
Watts, Mikey, and Yuchen Cui. 2026. "Rephrase Before You Act: Characterizing and Mitigating Language Sensitivity in Vision-Language-Action Models." https://omanscience.com/ar/articles/rephrase-before-you-act-characterizing-and-mitigating-language-sensitivity-in-vision-language-action-models.
هارفارد
Watts, M. and Cui, Y. (2026) 'Rephrase Before You Act: Characterizing and Mitigating Language Sensitivity in Vision-Language-Action Models', Available at: https://omanscience.com/ar/articles/rephrase-before-you-act-characterizing-and-mitigating-language-sensitivity-in-vision-language-action-models.
فانكوفر
Watts M, Cui Y. Rephrase Before You Act: Characterizing and Mitigating Language Sensitivity in Vision-Language-Action Models. https://omanscience.com/ar/articles/rephrase-before-you-act-characterizing-and-mitigating-language-sensitivity-in-vision-language-action-models
IEEE
M. Watts, and Y. Cui, "Rephrase Before You Act: Characterizing and Mitigating Language Sensitivity in Vision-Language-Action Models," https://omanscience.com/ar/articles/rephrase-before-you-act-characterizing-and-mitigating-language-sensitivity-in-vision-language-action-models.