VSpark accelerates autoregressive vision-language-action models by improving speculative decoding with calibrated action-noise rollouts, hidden-feature alignment, and reuse of rejected-draft continuations, achieving lossless 1.28x–2.10x speedups across multiple robot manipulation tasks.
Our 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…
autoregressive vision-language-action modelsVLAsspeculative decodingdraft modellossless decodingcalibrated action noisefeature alignmenthidden featurescontinuation reuseaction tokenizers
- Upvotes
- 7
- Authors
- 6
- Published
- Nov 17, 2026
Summary and abstract excerpt via Hugging Face Papers ↗ · arXiv.