titleVSpark: Speculative Decoding for Autoregressive Vision-Language-Action Models
authorsSuho Yoo, Youngjoon Jang, Hyebin Cho, Taeyeop Lee, Jaewook Jeong, Joon Son Chung
date2026-11-17
venuearXiv (preprint, not peer reviewed)
idarXiv:2611.22666
affiliationKAIST AI
upvotes5 (Hugging Face Daily Papers, 2026-11-21)
linksabs pdf hf code
cite
primerVision-language-action models
relatedOmniShowcase: Omnidirectional Subject-Consistent Video Generation from Multi-View Images
Look Only When Needed: Certificate-Guided Visual Routing for Efficient Diffusion VLMs
Language Models Can Control Their Own Attention
OmniRetrieval: Unified Retrieval across Heterogeneous Knowledge Sources
GRACE: Generation-aware latent compression for efficient video generation
abstractOur aim is to accelerate autoregressive vision-language-action models (VLAs), which generate robot actions one token at a time. Speculative decoding makes this possible: a lightweight draft model proposes several tokens at once, and the VLA verifies them in a single pass, accepting only those that match its own predictions, so decoding remains lossless. The speedup, however, is limited in robot control, where many different tokens encode nearly the same motion and a draft close to the correct action is still rejected. To address this, our approach, VSpark, is built upon three design choices: (i) rollouts collected with calibrated action noise, which expose the draft to a wider range of robot states; (ii) feature alignment, which trains the draft to match the VLA's hidden features rather than its output tokens alone; and (iii) continuation reuse, which keeps the predictions the VLA has already computed along rejected drafts as candidates for the next verification step, instead of discarding them. On four VLAs with different action tokenizers and four manipulation benchmarks, VSpark achieves 1.28x to 2.10x speedup while keeping the VLA's decoding lossless.

bibtex

@misc{yoo2026vspark,
  title={VSpark: Speculative Decoding for Autoregressive Vision-Language-Action Models},
  author={Suho Yoo and Youngjoon Jang and Hyebin Cho and Taeyeop Lee and Jaewook Jeong and Joon Son Chung},
  year={2026},
  eprint={2611.22666},
  archivePrefix={arXiv},
  url={https://arxiv.org/abs/2611.22666}
}