# FastVLA：面向高效视觉-语言-动作模型的任务条件化与时间自适应视觉 Token 压缩

> 原文链接：https://robotics.hellobike.com/papers/2026-10-15-2610-22146
> 频道：papers | 更新日期：2026-10-15 | 标签：arxiv, cs.RO
> 核心摘要：FastVLA以任务注意力和时间自适应压缩视觉 Token，在仿真与Franka真机上提升VLA推理效率并保持成功率。

# FastVLA：面向高效视觉-语言-动作模型的任务条件化与时间自适应视觉 Token 压缩

::: v-pre
> 原文链接：[FastVLA: Task-Conditioned and Temporally Adaptive Visual Token Compression for Efficient Vision-Language-Action Models](https://arxiv.org/abs/2610.22146)（arXiv 2610.22146 · 发布日期：2026-10-15）

**FastVLA: Task-Conditioned and Temporally Adaptive Visual Token Compression for Efficient Vision-Language-Action Models**

FastVLA以任务注意力和时间自适应压缩视觉 Token，在仿真与Franka真机上提升VLA推理效率并保持成功率。

## 核心要点

- VLA闭环部署受推理效率限制，大量视觉 Token 主导Transformer计算与延迟。
- FastVLA免训练，用任务到视觉注意力评分，并以逆Simpson指标自适应设置压缩率。
- 仿真跨LIBERO和SIMPLER评估，降低LLM骨干FLOPs与动作延迟并保持任务成功。
- 真机Franka Research 3端到端延迟降低20%、成功率相当，代码已在GitHub开源。

**验证范围与局限**：验证范围包括LIBERO、SIMPLER仿真和一台Franka Research 3真机。

## 英文摘要 (Abstract)

Vision-Language-Action (VLA) models present a promising paradigm for general-purpose robot manipulation, but closed-loop deployment is often limited by inference efficiency: at each timestep, high-resolution, often multi-view images are encoded into hundreds of visual tokens that dominate transformer computation and latency. Token compression via attention is a natural solution but is challenging for VLA control because (1) decision-critical tokens are determined by the task, not generic visual saliency, and (2) the token budget varies substantially across steps, so static metrics and budgets waste computation or discard control-critical tokens. We introduce FastVLA, a training-free, task-aware, and temporally adaptive visual token compression framework. At each step, FastVLA scores token importance via task-to-vision attention and adaptively sets the compression rate from the importance-distribution uniformity, measured by an inverse-Simpson metric with online sliding-window calibration. We evaluate FastVLA on OpenVLA, OpenVLA-OFT, and CogACT across LIBERO and SIMPLER, and on a physical Franka Research 3 robot. In simulation, it reduces LLM-backbone FLOPs by 35-59\% and action latency by 9-32\% (a 1.09$\times$-1.46$\times$ speedup) while maintaining task success, achieving a comparable or better efficiency-performance trade-off than existing plug-and-play token compression/pruning/reuse baselines. On the real robot, it cuts end-to-end latency by 20\% (1.25$\times$ speedup) at comparable success. Code is available at https://github.com/jiuyixu25/FastVLA.

作者：Jiuyi Xu, Chengyuan Qian, Zhengzhong Tu, Yang Sui, Yangming Shi
:::

## 原始参考来源 (Sources)

- [arXiv 2610.22146：FastVLA: Task-Conditioned and Temporally Adaptive Visual Token Compression for Efficient Vision-Language-Action Models](https://arxiv.org/abs/2610.22146)