InternW0：用于高效真实世界交互的基础物理世界模型
InternW0: A Foundational Physical World Model for Efficient Real-World Interactions
浏览论文内容
中文总结 AI 辅助
InternW0是上海AI实验室首个物理世界模型，通过非对称视频-动作架构和流匹配联合学习视觉动态与机器人控制，在7,200小时数据上训练，支持科学任务，实现高效真实世界交互。
中文摘要 AI 辅助
物理智能需要的不仅仅是预测世界如何演变：预测必须在世界持续变化时保持可执行性。我们推出了InternW0，这是上海人工智能实验室InternW物理世界模型系列的首个实例，其构建围绕全模态接口、异步多频率处理以及在部分观测和外部影响下的局部物理建模。InternW0通过一种带有流匹配的非对称视频-动作架构，联合学习未来视觉动态和连续机器人控制。一个高容量视频专家提供更长视界的预测上下文，而一个轻量级动作专家以更快的时标运行。InternW0不是为每次动作更新重新生成未来，而是重用逐层的K/V，并通过观测条件上下文路由使其适应新观测到的状态。领域特定接口和软提示支持异构具身，而接触感知的后训练结合力和触觉信号用于接触丰富的操作。我们在约7,200小时的异构机器人和自我中心数据上训练InternW0，包括EgoLab，一个275小时的真实实验室自我中心数据集。评估涵盖仿真基准和真实世界科学任务，包括一个15阶段的金属-有机框架合成流程和5阶段的接触与力感知灵巧操作，用于通用定量移液。这些结果推进了可扩展、异步且科学原生的物理世界模型，以实现通用且高效的真实世界交互。
英文摘要
Physical intelligence requires more than predicting how the world may evolve: predictions must remain actionable as the world continues to change. We introduce InternW0, the first instantiation of the InternW physical world model series from Shanghai AI Laboratory, built around omnimodal interfaces, asynchronous multi-frequency processing, and local physical modeling under partial observations and external influences. InternW0 jointly learns future visual dynamics and continuous robot control through an asymmetric video--action architecture with flow matching. A high-capacity video expert provides longer-horizon predictive context, while a lightweight action expert operates at a faster timescale. Instead of regenerating the future for every action update, InternW0 reuses layerwise K/V and adapts it to newly observed states through observation-conditioned context routing. Domain-specific interfaces and soft prompts support heterogeneous embodiments, while contact-aware post-training incorporates force and tactile signals for contact-rich manipulation. We train InternW0 on approximately 7,200 hours of heterogeneous robot and egocentric data, including EgoLab, a 275-hour real-laboratory egocentric dataset. Evaluation spans simulation benchmarks and real-world scientific tasks, including a 15-stage metal--organic framework synthesis workflow and 5-stage contact- and force-aware dexterous manipulation for general-purpose quantitative pipetting. These results advance scalable, asynchronous, and science-native physical world models for universal and efficient real-world interactions.
发表机构
- Shanghai AI Laboratory（上海人工智能实验室）
机构由 AI 辅助整理，请以论文原文为准。
