衡宇 发自 凹非寺 量子位 | 公众号 QbitAI 一个 （暂时） 只做具身大脑的公司，抛出了一个 只有2.4B参数 的具身模型。 目前行业风向标如Physical Intelligence的π 0总计约33亿参数，π 0.6的参数量也约莫在50亿以上。 在一个甚至连硬件形态都还没定型的行业里，2.4B参数到底够不够用？ 这家公司给出的答案是，够用。 而且 足以支撑它实时处理三视角的728x728画面，推理延迟仅60毫秒；配合强化学习机制，它还能在真机上不断试错进化。 这就是具身智能创企原力灵机推出的首个具身原生模型产品DM0。 2.4B的轻量小蛋糕， RTX 5090就能跑。 因为从零训练以及对具身数采有不同于行业的看法等原因，该公司称它为"首个具身原生大模型"。 与模型同时发布的还有开源具身原生框架Dexbotic 2.0，以及具身原生量产工作流 DFOL。 这具身软件三件套背后技术路线的操盘手，是原力灵机合伙人、负责大模型的周而进。 他在AI圈早已名声在外。 周而进现在才33岁，但 这人已经在AI领域出名13年了 —— 早在2013年，深度学习和人工智能还是冷门的时候，大二的旷视实习生周而进就以一作身 ...

作者｜ Stone Jin 微信公众号｜ipozaozhidao 据IPO早知道消息，原力灵机于2月10日举行首届"具身原生——原力灵机技术开放日"，这也是原力 灵机成立后的首次集中亮相。 原力灵机联合创始人&CEO唐文斌表示：" 2026年不是具身智能的元年，而是'具身原生'的元 年。 我们需要的不仅 仅是一个'能在机器人上运行的大模型'，而是一个'智能本质和形成机制都根植于物 理交互的新AI范式'，我们将其命名为'具身原生'。 " 唐文斌进一步指出，具身原生（Embodied Native）是机器人规模化、场景化应用的支点， 核心理 念聚焦机器人真实应用场景，从0开始进行数据定义，模型训练，应用部署，最大程度统一通用交 互，生成广义动作，实现机器人与真实环境的原生融合，驱动具身智能范式跃迁。 值得一提的是， 在本次发布会上，原力灵机发布了全球首个从零开始训练的具身原生大模型DM0， 其充分体现了"多源数据预训练"、"多任务、跨机型预训练"与"空间推理思维链"三大核心特征。 DM0最显著的突破在于其 "智能密度" 。其2.4B参数版本，在权威真机评测基准 RoboChallenge 上，一举夺得单任务与多任 ...

21世纪经济报道新质生产力研究院综合报道 早上好，新的一天又开始了。在过去的24小时内，科技行业发生了哪些有意思的事情？来跟21tech一起 看看吧。 【巨头风向标】 近期，字节跳动最新视频生成模型Seedance2.0在即梦、豆包、小云雀等产品开启内测，引发国内国外 广泛关注。值得注意的是，由于该模型在真实感上的突出表现，引发不少网友对可能混淆现实的担忧。 对于用户反馈，字节跳动很快做出反应。2月9日，即梦创作者社群中，平台运营人员发布消息 称："Seedance2.0在内测期间收获了远超预期的关注，感谢大家的使用反馈。为了保障创作环境的健康 可持续，我们正在针对反馈进行紧急优化，目前暂不支持输入真人图片或视频作为主体参考"，并表示 平台深知创意的边界是尊重，产品调整后会以更完善的面貌与大家正式见面。 追觅创始人俞浩隔空喊话余承东：在哪上班不是上？要不加入追觅吧 2月10日，追觅科技创始人在其个人微博发文@华为常务董事、产品投资评审委员会主任、终端BG董事 长，俞浩表示"在哪上班不是上？要不加入追觅吧！"在评论区，有网友调侃"俞总调戏余总"，俞浩回 复"真诚邀约"，并且俞浩还在评论区提到"你来追觅，我不会对 ...

Core Insights - The company, Yuanli Lingji, has launched three core products: the first embodied native large model DM0, the embodied native development framework Dexbotic 2.0, and the embodied native application mass production workflow DFOL, emphasizing that 2026 will be the year of embodied natives rather than just embodied intelligence [1][3] Product Launch - DM0 is the world's first embodied native large model, designed to operate in complex environments and complete human tasks accurately from its inception, integrating multimodal internet information and unique embodied scene data such as driving behavior and robot operations [3] - Dexbotic 2.0 features a modular architecture that allows developers to build their embodied applications in a Lego-like manner, offering five core advantages over its predecessor, including independent upgrades and replacements of components [3][4] - DFOL introduces a data feedback mechanism that enables continuous evolution of the system through a closed loop of cloud training, on-site execution, data feedback, and model updates, enhancing flexibility and adaptability in real-world environments [4][5] Strategic Collaborations - The company has partnered with prestigious institutions like Tsinghua University and Princeton to create a unified infrastructure for embodied intelligence, similar to what PyTorch has done for deep learning, aiming to lower development barriers and foster innovation [4]

Core Viewpoint - The article discusses the significant advancements in embodied intelligence, particularly through the launch of the Dexbotic 2.0 framework and its collaboration with RLinf, marking a pivotal moment in the industry towards a "native embodied" era of AI [3][5][9]. Group 1: Framework and Collaboration - The Dexbotic 2.0 framework aims to standardize the infrastructure for embodied intelligence, similar to how PyTorch revolutionized deep learning [5][16]. - The collaboration with Tsinghua University and RLinf focuses on enhancing the capabilities of embodied AI through a unified framework that integrates perception, decision-making, and execution [3][5][19]. - The introduction of the DM0 model and the DFOL workflow signifies a comprehensive approach to developing and deploying embodied applications [6][51]. Group 2: Embodied Native Concept - "Embodied Native" is defined as a concept that emphasizes a closed-loop system of perception, decision-making, and execution, allowing AI to interact with the physical world effectively [15][13]. - The framework promotes the use of real-world data and multi-modal training to enhance the model's understanding and interaction with its environment [17][41]. - The transition from a "big model brain + mechanical limbs" approach to a fully integrated embodied system is highlighted as a key evolution in the field [12][13]. Group 3: Technical Innovations - Dexbotic 2.0 features a modular design that maintains high flexibility while ensuring end-to-end processing, allowing for independent upgrades of perception, cognition, and control modules [21][33]. - The framework integrates various models and capabilities, including visual-language-action (VLA) and navigation, to achieve comprehensive task execution [37][38]. - The introduction of a standardized data format (Dexdata) and a unified training pipeline addresses the fragmentation in the development of embodied intelligence [45][46]. Group 4: Performance and Evaluation - The DM0 model, with 2.4 billion parameters, has achieved high performance in real-world evaluations, demonstrating its capability in both single and multi-task scenarios [57][58]. - The RoboChallenge benchmark is established to provide a fair evaluation of embodied models, ensuring that performance metrics reflect true capabilities rather than optimized scores [46][57]. - The DFOL workflow enables continuous improvement of robotic systems through real-time data feedback, enhancing their operational efficiency [62][65]. Group 5: Future Insights - The article emphasizes the importance of integrating multi-modal sensory inputs, such as touch and auditory capabilities, to enhance the modeling of the physical world [74]. - The rapid evolution of embodied intelligence is noted, with expectations for significant advancements in the near future, akin to the pace seen in large model developments [73][75]. - The company advocates for an open-source approach to foster collaboration and innovation within the embodied intelligence community, aiming to lower barriers for developers [68][71].