ScaleVLA: Scaling Vision-Language-Action Models to 552 Billion Parameters
ScaleVLA: Scaling Vision-Language-Action Models to 552 Billion Parameters
ScaleVLA connects the 552B DeepSeek-V4.1-Flash backbone with trainable action DiT, comparing VLMs from 2B to 552B under shared training settings. 10 demonstrations surpass 2B backbone with 50 training samples (87.8% vs. 85.6%), LIBERO-PRO success rate reaches 45.1% (highest 31.3% for small models), charger insertion success rate improves from 4/10 to 9/10.
Published on Nov 16
Authors:
,
,
,
,
,
,
,
,
,
Abstract
Frontier Vision-Language Models (VLMs) such as GPT-6 Astra have demonstrated remarkable capabilities in robot control, suggesting that broad pretraining provides knowledge valuable for physical action. Yet most Vision-Language-Action (VLA) policies still rely on much smaller VLM backbones, leaving this potential largely untapped. We ask whether frontier-scale representations can provide a stronger foundation for robot learning. We introduce ScaleVLA, which connects a 552B DeepSeek-V4.1-Flash backbone to a trainable action DiT, and systematically compare VLMs from 2B to 552B under a shared training setup. We evaluate three fundamental capabilities of robot foundation models. In sample efficiency, just 10 demonstrations outperform a 2B backbone trained with 50 (87.8% vs. 85.6%). In generalization, LIBERO-PRO success reaches 45.1%, compared with at most 31.3% for smaller backbones. In precise manipulation, real-world charger insertion improves from 4/10 to 9/10 successes, with gains concentrated in contact-sensitive alignment. Despite its scale, ScaleVLA achieves 216 ms median latency per action chunk. Our findings suggest that frontier VLMs can serve not only as reasoning agents, but also as powerful representational foundations for efficient and generalizable robot policies.
View arXiv page View PDF Project page Add to collection
Get this paper in your agent:
hf papers read 2611.21607
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash
Models citing this paper 0
No model linking this paper
Cite arxiv.org/abs/2611.21607 in a model README.md to link it from this page.
Datasets citing this paper 0
No dataset linking this paper
Cite arxiv.org/abs/2611.21607 in a dataset README.md to link it from this page.
Spaces citing this paper 0
No Space linking this paper
Cite arxiv.org/abs/2611.21607 in a Space README.md to link it from this page.
Collections including this paper 1
Source: Hugging Face Daily Papers · huggingface.co