| title | VisualThink-VLA: Visual Intermediate Reasoning for Effective and Low-Latency Vision-Language-Action Policies |
|---|---|
| authors | Mingjian Gao, Wenqiao Zhang, Yuqian Yuan, Yang Dai, Binhe Yu, Zheqi Lv, Haoyu Zheng, Jiaqi Zhu, Zhiqi Ge, Zixuan Wan, Siliang Tang, Yueting Zhuang |
| date | 2026-05-28 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2605.30011 |
| affiliation | ZJU4EmbodiedAI |
| upvotes | 3 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf code |
| primer | Vision-language-action models |
| abstract | Recent work has begun to equip vision-language-action (VLA) policies with explicit intermediate reasoning. In embodied control, however, textual chain-of-thought is a poor fit: irrelevant or weakly textual information can interfere with action prediction, while autoregressive text decoding adds too much latency for real-time closed-loop execution. We present VISUALTHINK-VLA, a visual intermediate-reasoning framework for accurate, low-latency VLA policies. Our bootstrapping philosophy is to guide action with effective visual thinking: VISUALTHINK-VLA bootstraps action prediction through a compact visual-evidence interface that preserves spatial precision while avoiding decoding overhead. Besides, to further improve performance and efficiency, VISUALTHINK-VLA adopts a tailored selective routing mechanism to learn the visual evidence tokens, enabling low-latency inference while preserving high-capacity specialization. We also introduce VisualEvidence-Kit, a supervision-and-audit resource centered on a VisualEvidence-Agent that constructs a 754.7k VLA instructions VisualEvidence-Set for route supervision and counterfactual faithfulness tests. Across multiple benchmarks and real-robot evaluation, VISUALTHINK-VLA achieves the highest success rate on most benchmarks while reducing the multi-second latency of reasoning-augmented baselines to the sub-second regime. For example, on BridgeData V2, it reduces step latency from 8.377,s with ECoT to 0.367,s, achieving a 22.8 times speedup. |
cite
@misc{gao2026visualthink,
title={VisualThink-VLA: Visual Intermediate Reasoning for Effective and Low-Latency Vision-Language-Action Policies},
author={Mingjian Gao and Wenqiao Zhang and Yuqian Yuan and Yang Dai and Binhe Yu and Zheqi Lv and Haoyu Zheng and Jiaqi Zhu and Zhiqi Ge and Zixuan Wan and Siliang Tang and Yueting Zhuang},
year={2026},
eprint={2605.30011},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2605.30011}
}