At its core, robotic manipulation is a problem of vision-to-geometry mapping ($f(v) \rightarrow G$). Physical actions are fundamentally defined by geometric properties like 3D positions and spatial relationships. Consequently, we argue that the foundation for generalizable robotic control should be a vision-geometry backbone, rather than the widely adopted vision-language or video models. Conventional VLA and video-predictive models rely on backbones pretrained on large-scale 2D image-text or temporal pixel data. While effective, their representations are largely shaped by semantic concepts or 2D priors, which do not intrinsically align with the precise 3D geometric nature required for physical manipulation. Driven by this insight, we propose the Vision-Geometry-Action (VGA) model, which directly conditions action generation on pretrained native 3D representations. Specifically, VGA replaces conventional language or video backbones with a pretrained 3D world model, establishing a seamless vision-to-geometry mapping that translates visual inputs directly into physical actions. To further enhance geometric consistency, we introduce a Progressive Volumetric Modulation module and adopt a joint training strategy. Extensive experiments validate the effectiveness of our approach. In simulation benchmarks, VGA outperforms top-tier VLA baselines including $π_{0.5}$ and GeoVLA, demonstrating its superiority in precise manipulation. More importantly, VGA exhibits remarkable zero-shot generalization to unseen viewpoints in real-world deployments, consistently outperforming $π_{0.5}$. These results highlight that operating on native 3D representations-rather than translating through language or 2D video priors-is a highly promising direction for achieving generalizable physical intelligence.

翻译：核心而言，机器人操控是一个视觉到几何映射($f(v) \rightarrow G$)的问题。物理动作从根本上由三维位置与空间关系等几何属性所定义。因此，我们认为通用机器人控制的基础应是视觉-几何主干网络，而非当前广泛采用的视觉-语言或视频模型。传统视觉-语言动作(VLA)模型与视频预测模型依赖在规模化二维图像-文本数据或时序像素数据上预训练的主干网络。尽管这些模型效果显著，但其表征主要由语义概念或二维先验塑造，本质上与物理操控所需的三维几何精确性不兼容。基于这一洞见，我们提出视觉-几何-动作(VGA)模型，该模型直接基于预训练原生三维表征生成动作。具体而言，VGA用预训练三维世界模型替代传统语言或视频主干网络，构建起将视觉输入直接转化为物理动作的无缝视觉到几何映射。为进一步增强几何一致性，我们引入渐进式体积调制模块，并采用联合训练策略。大量实验验证了本方法的有效性。在仿真基准测试中，VGA表现优于包括$π_{0.5}$和GeoVLA在内的顶级VLA基线模型，展现了其在精细操控中的优越性。更重要的是，VGA在真实场景部署中对未见过视角展现出显著的零样本泛化能力，持续优于$π_{0.5}$。这些结果凸显了基于原生三维表征——而非通过语言或二维视频先验进行转译——是实现通用物理智能极具前景的方向。