# 1Step-WAM：使用 1-Step Action Generation 让你的 World Action Model 更快

> 原文链接：https://robotics.hellobike.com/papers/2026-10-11-2610-15625
> 频道：papers | 更新日期：2026-10-11 | 标签：arxiv, cs.RO
> 核心摘要：针对WAM迭代去噪与串行执行瓶颈，1Step-WAM用校准和受限解码实现一步生成，并在RoboTwin、LIBERO和真机上验证。

# 1Step-WAM：使用 1-Step Action Generation 让你的 World Action Model 更快

::: v-pre
> 原文链接：[1Step-WAM: Your World Action Model but Faster using 1-Step Action Generation](https://arxiv.org/abs/2610.15625)（arXiv 2610.15625 · 发布日期：2026-10-11）

**1Step-WAM: Your World Action Model but Faster using 1-Step Action Generation**

针对WAM迭代去噪与串行执行瓶颈，1Step-WAM用校准和受限解码实现一步生成，并在RoboTwin、LIBERO和真机上验证。

## 核心要点

- WAM统一视觉世界建模与动作生成，但真实部署受迭代去噪和串行执行限制。
- 方法利用去噪轨迹近似直线，用轻量校准器修正首步速度并约束动作解码。
- 一步动作生成可与视觉预填充并行，打破传统先想象后动作的顺序瓶颈。
- 在FastWAM、ImageWAM和Kairos上分别实现7.44×、5.98×和24.84×加速，代码已开源。

**验证范围与局限**：验证范围包括RoboTwin、LIBERO，以及使用FastWAM骨干的真实机器人部署。

## 英文摘要 (Abstract)

World Action Models (WAMs) unify visual world modeling and action generation, yet their real-world deployment is heavily bottlenecked by iterative action denoising and the sequential execution of visual prefilling and action prediction. We demonstrate that the action-denoising trajectories in WAMs are approximately straight, rendering the first-step velocity a strong estimate of the final action direction. Building on this observation, we introduce 1Step-WAM, a framework that transforms multi-step action denoising into one single forward. Specifically, we employ a lightweight calibrator to directly correct the first step velocity, and use bandwidth-constrained action decoding to further reduce errors in the calibrated one-step actions. Compressing action generation into a single step enables it to execute in parallel with visual prefilling, completely breaking the sequential bottleneck of the traditional imagination-then-action paradigm. Evaluations on RoboTwin and LIBERO show that 1Step-WAM largely preserves multi-step task success with only about 13.4M trainable parameters, while achieving inference speedups of $7.44\times$, $5.98\times$, and $24.84\times$ for FastWAM, ImageWAM and Kairos, respectively. Furthermore, real-world robotic deployments using the FastWAM backbone confirm a consistent 7.44$\times$ speedup while maintaining close success rate to the multi-step baseline. Code is avaliable at https://github.com/changdong1687/1Step-WAM.

作者：Rundong Su, Pengtao Chen, Haohan Geng, Sha Zhang, Zhihang Yuan, Yicheng Rong, Chong Yu, Hao Huang, Tao Chen
:::

## 原始参考来源 (Sources)

- [arXiv 2610.15625：1Step-WAM: Your World Action Model but Faster using 1-Step Action Generation](https://arxiv.org/abs/2610.15625)