OpenWAM: An Open, Modular Exploration Towards Systematic World-Action Model Pretraining

世界—动作模型从视频生成先验中继承世界知识，并通过具身经验将其转化为可执行的控制信号。然而，现有系统通常是单体式的：生成骨干、视觉表示、架构、信息流、推理过程和训练数据紧密耦合，使人难以判断哪些设计选择真正重要以及原因何在。作者提出 OpenWAM，一个把世界—动作预训练变成受控实验计划的开放研究栈。OpenWAM-Infra 将 WAM 设计空间分解为可组合模块，并统一训练、推理、部署和评测。在此基础上，OpenWAM-Study 通过受控实验研究三个问题：应继承什么、世界与动作学习如何交互、二者协同如何扩展，并总结出三项原则：上游知识需要通过能力充分的生成骨干和紧凑且信息丰富的潜空间进行迁移；世界—动作协同需要专用动作容量、显式世界到动作的信息流以及同步联合去噪；具身预训练主要改善分布外泛化，而在人类第一视角和机器人数据上进行单阶段共训，可以整合世界覆盖与动作落地。基于这些原则，作者构建 OpenWAM-α，在约 6400 小时的人类第一视角和机器人数据上预训练，并在仿真与真实世界基准中评测。八个仿真基准和真机实验覆盖单臂、双臂到灵巧手等多种形态，OpenWAM-α 从仿真到物理世界都保持了较强表现。作者开放完整技术栈，包括基础设施、评测协议、预训练模型和数据配方，以推动后续研究。

World-Action Models inherit world knowledge from video-generative priors, and channel it into executable control signals through embodied experience. Existing systems, however, are monolithic: the generative backbone, visual representation, architecture, information flow, inference procedure, and training data are tightly coupled, obscuring which design choices matter and why. We introduce OpenWAM, an open research stack that turns world-action pretraining into a controlled experimental program. OpenWAM-Infra factorizes the WAM design space into composable modules with unified training, inference, deployment, and evaluation. On this substrate, OpenWAM-Study examines three questions through controlled experiments: what to inherit, how world and action learning interact, and how their synergy scales; and distills three principles: upstream knowledge transfers through a sufficiently capable generative backbone and a compact, information-rich latent space; world-action synergy requires dedicated action capacity, explicit world-to-action information flow, and synchronized joint denoising; and embodied pretraining principally improves out-of-domain generalization, with one-stage co-training over egocentric and robot data integrating world coverage and action grounding. Composing these principles, we build OpenWAM-α, an open WAM pretrained on roughly 6,400 hours of egocentric human and robot data and evaluated across simulation and real-world benchmarks. Across the eight simulation benchmarks and the real-robot experiments, which together span embodiments from single-arm and bimanual manipulation to dexterous hands, OpenWAM-α delivers consistently excellent performance, sustaining its top-tier standing from simulation to the physical world. We release the full stack, including infrastructure, evaluation protocols, pretrained models, and data recipes, to facilitate future research.