[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-nvidia-cosmos-3-world-foundation-physical-ai":3,"topics-all":33,"news-related-f8f3fa67-16b3-43f0-8330-17b55846f4ec":52},{"id":4,"title":5,"summary":6,"content":6,"original_url":7,"source_id":8,"tags":9,"translations":20,"news_slug":26,"published_at":27,"created_at":28,"modified_at":29,"is_published":30,"publish_type":31,"image_url":13,"view_count":32},"f8f3fa67-16b3-43f0-8330-17b55846f4ec","NVIDIA Cosmos 3：开源世界基础模型能否让机器人真正「理解」物理世界？","5月31日，NVIDIA在台北COMPUTEX上发布了 Cosmos 3，这是首个将视觉推理、多模态生成与动作预测整合在一起的开放世界基础模型（World Foundation Model），被视为物理AI领域的重大突破。\n\n传统视频生成模型往往只是「生成」画面，而Cosmos 3的核心创新在于引入了推理模块——先用Mixture-of-Transformers架构理解当前场景，再基于上下文生成后续内容。这意味着模型不只是生成视频，而是真正在「思考」物理世界的因果关系。\n\n在 benchmark 表现上，Cosmos 3已登顶多个开放权重排行榜，包括 VANTAGE-Bench（智能城市与空间理解）、Physics-IQ、R-Bench 及 PAI-Bench 等物理推理相关评测。对于机器人开发者而言，这意味着可以通过模型从海量视频中合成符合物理规律的动作数据——这恰恰是当前机器人学习最稀缺的资源。\n\n物理AI的瓶颈从来不是感知，而是数据。Cosmos 3的价值不在于又刷新了多少榜单分数，而在于它提供了一种可扩展的合成数据生产方式。如果开源社区能在此基础上快速迭代，通用机器人的落地时间线可能会比我们想象的更近。","https:\u002F\u002Fblogs.nvidia.com\u002Fblog\u002Fcosmos-3-physical-ai-open-world-foundation-model\u002F","474eef8c-e0c3-46cf-adee-c089558220f9",[10,14,17],{"id":11,"name":12,"slug":12,"description":13,"color":13},"e676a5cf-1f24-472f-a765-86fa21a1bc3c","ai-model",null,{"id":15,"name":16,"slug":16,"description":13,"color":13},"8dac812d-3839-4abe-a855-5f56ec9515fd","nvidia",{"id":18,"name":19,"slug":19,"description":13,"color":13},"ebe5dcd1-46b1-4298-b8c2-8e0e2f456e56","video-generation",[21],{"id":22,"lang":23,"title":24,"summary":25,"content":13},"fdc3af4b-92ca-44db-9ac6-b852491f4a23","en","NVIDIA Cosmos 3: open world models meet physical robots","NVIDIA released Cosmos 3 on June 2, an open-source world foundation model for physical AI. The model is trained on millions of hours of physical-world video and is specifically designed for robotics, autonomous driving, and embodied AI training. The bet: a true \"world foundation model\" is the missing layer for Physical AI to break through.","nvidia-cosmos-3-world-foundation-physical-ai","2026-06-02T04:00:00Z","2026-06-02T04:05:15.202138Z","2026-08-19T02:08:40.142862Z",true,"agent",187,[34,43],{"slug":35,"tag_slug":35,"title_zh":36,"title_en":37,"intro_zh":38,"intro_en":39,"id":40,"is_active":30,"created_at":41,"modified_at":42},"ai-for-science","AI for Science 2026：从 UniPert 到 GPT-Rosalind 的硬核进化","AI for Science 2026: from UniPert to GPT-Rosalind","生命科学、化学材料、物理世界模型——AI 正在从\"语言工具\"变成\"实验伙伴\"。本专题收录 AI 在三大科学方向的关键节点：UniPert 统一基因与化学扰动空间、GPT-Rosalind 端到端生命科学推理、达摩院 AI 智能体 28 小时找到 4 种超导新材料、Anthropic Claude Science 把工作台做成标准品。","From language tool to lab partner — AI is reshaping life sciences, chemistry\u002Fmaterials, and physical world models. This topic covers the key milestones: UniPert unifying genetic-chemical perturbation spaces, GPT-Rosalind's end-to-end life-sciences reasoning, DAMO's AI agent discovering 4 superconducting materials in 28 hours, and Anthropic's Claude Science workbench going mainstream.","988a4300-5fab-41c4-b5d8-63711a2dc757","2026-09-10T01:34:15.296649Z","2026-09-10T01:34:15.296663Z",{"slug":44,"tag_slug":44,"title_zh":45,"title_en":46,"intro_zh":47,"intro_en":48,"id":49,"is_active":30,"created_at":50,"modified_at":51},"h3-series","MiniMax H3 系列：从开源权重到 35 倍吞吐","MiniMax H3 Series: from open weights to 35x throughput","MiniMax H3 自 2026 年 8 月开源以来节奏密集：官方把生成、参考与编辑收回一个模型；ComfyUI 当天压进 RTX 3060；摩尔线程 3 小时完成国产 GPU 适配；fal 后训练版把吞吐拉到 35 倍；FastH3 蒸馏再砍推理成本。本专题持续追踪 H3 的发布—开源—蒸馏—部署全链路。","Since MiniMax open-sourced H3 in August 2026 the pace has been relentless: one unified omni-modal model, same-day ComfyUI support down to an RTX 3060, a 3-hour Day-0 port to Moore Threads GPUs, fal's post-trained H3 Max at 35x throughput, and FastH3 distillation cutting inference cost further. This topic tracks the full H3 chain — release, open weights, distillation, deployment.","83ef0daa-3c31-4cb3-86ed-e5ee58654d5f","2026-09-08T07:33:19.942193Z","2026-09-08T07:33:19.942209Z",{"items":53},[54,59,64,69,74,79],{"id":55,"title":56,"news_slug":57,"published_at":58},"03d88e2a-526c-4c39-b9b6-e90ac0ff65e0","PixelUMM 拆掉 VAE:NVIDIA 用 8B 单 Transformer 在像素空间统一图文视频","pixelumm-encoder-free-unified-multimodal-8b","2026-09-29T21:59:00+00:00",{"id":60,"title":61,"news_slug":62,"published_at":63},"75d2f385-00ef-4074-80bf-ee47ba05a4a4","NVIDIA × HF：Diffusers 微调上 H100，Wan 2.2\u002FFLUX.2 打通","nvidia-huggingface-diffusers-h100","2026-07-17T14:00:00+00:00",{"id":65,"title":66,"news_slug":67,"published_at":68},"b446777f-9949-47e4-8ca8-2d0fa7f14126","NVIDIA Cosmos 3 Edge 4B开源世界模型：物理AI实时推理从云端搬到产线","nvidia-cosmos-3-edge-4b","2026-07-16T14:00:00+00:00",{"id":70,"title":71,"news_slug":72,"published_at":73},"58764a46-a635-4f79-86c6-eefff0dea639","Flex-Forcing：NVIDIA 用 chunk 机制把 AR 与双向视频扩散塞进同一模型","nvidia-flex-forcing-ar-bidir-video","2026-07-08T18:24:00+00:00",{"id":75,"title":76,"news_slug":77,"published_at":78},"9652a175-6f95-461d-b9d4-946b44d0fccf","NVIDIA × Hugging Face：GR00T 1.7 接入 LeRobot","nvidia-hf-isaac-gr00t-cosmos","2026-07-07T06:00:26+00:00",{"id":80,"title":81,"news_slug":82,"published_at":83},"a2e8ac5b-ca51-4ddb-88d4-54373d1f0774","SUNTA 用\"惊奇度\"切分视频预测:东京大学让模型在 250 步后仍不崩溃","sunta-surprise-chunking-video","2026-07-04T16:00:00+00:00"]