Core Insights - DeepSeek has released an updated version of its R1 paper, adding 64 pages of technical details, significantly enhancing the original content [2][5][56] - The new version emphasizes the implementation details and training processes of the R1 model, showcasing a systematic approach to its development [10][11][17] Summary by Sections Paper Updates - The updated paper has expanded from 22 pages to 86 pages, providing a wealth of new information that resembles a textbook [3][6] - The revisions include a comprehensive breakdown of the R1 training process, which is divided into four main steps: cold start, inference-guided reinforcement learning, rejection sampling and fine-tuning, and alignment-guided reinforcement learning [13][14][15][16] Model Performance and Safety - The R1 model has shown a significant increase in reasoning capabilities, with a reported 5 to 7 times increase in the occurrence of reflective vocabulary as training progresses [21][22] - DeepSeek has implemented a safety control system that includes a dataset of 106,000 prompts to evaluate and enhance the model's safety, using a point-wise training method for the safety reward model [26][29] - The introduction of the risk control system has led to a notable improvement in the model's safety performance, with R1 achieving benchmark scores comparable to leading models [32][33] Team Stability and Industry Context - The core team behind the R1 paper has remained stable, with 18 key contributors still part of DeepSeek, indicating a low turnover rate in contrast to industry trends [41][47] - The article contrasts DeepSeek's team retention with the challenges faced by other companies in the AI sector, highlighting a more cohesive internal culture [48][49]

ProEdit团队 投稿 量子位 | 公众号 QbitAI 想给照片里的猫换个颜色，结果总是编辑失败？想让视频里的人换件衣服，人脸却糊成一片或完全改变？ 近日，来自中山大学iSEE实验室、香港中文大学MM Lab、新加坡南洋理工大学、香港大学的研究团队发布了最新研究成果 ProEdit 。 该方法通过对注意力机制和初始噪声潜在分布的"精准手术"，实现了超高精度的图像与视频编辑，且完全无需训练、即插即用。 △ 图1. ProEdit在图像和视频编辑上与现有方法的对比 为什么AI编辑总是"改不动"？ 目前，基于反演 （Inversion-based） 的编辑方法 （如RF-Solver、FireFlow） 通常采用全局注入策略： 为了保持背景尽量一致，它们 会将原图的大量信息强行"塞"进生成过程 。 在AI视觉编辑领域，如何在修改目标属性的同时，精准保留背景和非编辑属性的一致性，一直是个"鱼和熊掌"的难题。 但研究团队通过文本与图像的注意力可视化发现，这种做法存在严重的 "源图像信息过度注入" 问题： 注意力过度注入： 现有方法通过全局注入了过多的源图像注意力特征，导致模型更听源图像的话，而忽略了用户的编辑指令 ...

嘻疯 发自 凹非寺 量子位 | 公众号 QbitAI 国产具身智能基座模型，再次突破！ RoboChallenge真机评测榜单上，来自 自变 量机器人的 端到端具身智能基础模型WALL-OSS ，以46.43分的成绩，超越美国具身智能明星 公司Physical Intelligence的pi0 （π0） ， 总分 排名 全球第二 。 | | Beta | Home | Challenges | Runs | Leaderboard | News | Community | Eval Your Policy | Log In | | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | | | | | | Leaderboard | | | | | | | all tasks | table-30 > | Search by tag v | | Search by task v | | Is multitask > | | Search by model or user | Q | | Rank | Model/User | | Is multi ...

岗位均为全职，工作地点：北京中关村。 AI热潮还在汹涌，但如果你还不知道如何参与……那为什么不来 量子位 呢？ 我们是一家以 追踪AI新进展 为核心的内容平台，经过8年积累，目前拥有顶流影响力，广泛且备受认可的产业资源，以及时代风口的最佳观 测和学习生态位。 目前，我们有 三大方向 岗位招聘，希望你是 （或者能成为） 这三个方向的内容专家： 岗位面向： 加入我们，你可以获得： 编辑部 发自 凹非寺 量子位 | 公众号 QbitAI 以下是岗位详情： 所有岗位不同能力层级职位均在开放，欢迎结合个人履历和经验申请。 AI产业方向 岗位职责： AI产业方向 ：关注基建层创新，包含芯片、AI Infra、云计算； AI财经方向 ：关注AI领域创投和财报，跟踪产业链资本动向； AI产品方向 ：关注AI在应用和硬件终端方向的进展。 社招：覆盖编辑、主笔、主编各个层级，按能力匹配岗位； 校招：应届毕业生，接受实习且可转正。 站在AI浪潮之巅 ：第一时间接触和了解AI领域最新技术和产品，构建完整的AI认知体系。 玩转AI新工具 ：将各种AI新技术、新工具应用于工作，提升工作效率和创造力。 打造个人影响力 ：通过撰写独家原创内 ...

当通用能力主要通过大规模预训练获得之后，下一阶段的关键在于让已经具备通用能力的模型，在真实部署环境中持续进化。 这是智元机器人首席科学家 罗剑岚 博士在接受量子位采访时给出的论断。 智元机器人 投稿 量子位 | 公众号 QbitAI 2025年机器人领域最火的VLA让机器人通过预训练具备了相当的通用性，但与此同时，机器人能否长时间，稳定，高效地完成任务仍是一 个问号。 基于此，当机器人走出实验室，走向开放、复杂且持续变化的真实世界时，一个更核心的问题随之出现：如何真正实现通用机器人的规模化 部署与智能化运行。 为此，智元机器人具身研究中心提出 SOP（ScalableOnlinePost-training） ——一套面向真实世界部署的 在线后训练系统 。 这是业界首次在物理世界的VLA后训练中， 系统性地融合在线学习、分布式架构与多任务通才性 ，使机器人集群能够在真实环境中持续进 化，让个体经验在群体中高效复用，从而将"规模"转化为"智能"。 真实世界中的规模化智能增长挑战 要在真实世界中大规模运行，通用机器人必须同时满足两个看似矛盾的要求： 现有VLA预训练模型已经提供了强大的通用性。但 真实世界的部署受困 ...

Core Insights - The article discusses the emergence of numerous keywords in the AI product sector by 2025, highlighting transformative AI products that are leading the market [4] - The "AI 100" list by Quantum Bit Think Tank aims to evaluate and recognize the top AI products in China, reflecting the industry's evolution and future trends [4][12] Group 1: AI 100 List Overview - The "AI 100" list is divided into three main categories: "Flagship AI 100," "Innovative AI 100," and the top three products in ten popular sub-sectors [6] - The "Flagship AI 100" will focus on the strongest AI products of 2025, showcasing those that have achieved significant technological breakthroughs and practical application value [7] - The "Innovative AI 100" aims to identify products that are emerging in 2025 and have the potential to lead industry changes in 2026 [8] Group 2: Sub-sector Focus - The ten hottest sub-sectors for the top three products include AI browsers, AI agents, AI smart assistants, AI workstations, AI creation, AI education, AI healthcare, AI entertainment, Vibe Coding, and AI consumer hardware [9] Group 3: Application and Evaluation Criteria - The evaluation of the "AI 100" list employs a dual assessment system combining quantitative and qualitative measures, focusing on user data and expert evaluations [13] - Quantitative metrics include user scale, growth, activity, and retention, while qualitative assessments consider long-term potential, technology, market space, and user experience [13]

Core Viewpoint - The article highlights the successful IPO of Zhiyu, referred to as the "first stock of global large models," marking a significant milestone for Chinese AGI companies in the international capital market [1][36]. Group 1: IPO Details - Zhiyu officially listed on the Hong Kong Stock Exchange with the stock code 2513, opening at 120 HKD per share and achieving a market capitalization of over 52.8 billion HKD [2][3]. - The IPO raised over 4.3 billion HKD, with a subscription rate of 1159.46 times for the public offering and 15.28 times for the international offering [5][6]. - The company attracted a star-studded lineup of cornerstone investors, including major state-owned enterprises and international institutions, securing 29.8 billion HKD in subscriptions [10]. Group 2: Technological Strength - Zhiyu's flagship model, GLM-4.7, has achieved top rankings in various global AI benchmarks, showcasing its competitive edge against international models [15][18]. - The GLM architecture is compatible with over 40 domestic chipsets, and its AutoGLM 2.0 can control 80 million devices, with a daily token usage of 4.6 trillion [18][21]. - The company has established partnerships with over 50 international platforms, integrating its GLM model as a core capability [19][20]. Group 3: Financial Performance - Zhiyu's revenue has shown impressive growth, with a compound annual growth rate of 130%, increasing from 57.4 million HKD in 2022 to 312.4 million HKD in 2024, and a 325% year-on-year increase in the first half of 2025 [22][28]. - The company has adopted a MaaS model, with over 2.7 million enterprises and developers using its platform, and its subscription product has surpassed 100 million HKD in annual recurring revenue [25][27]. Group 4: Investment in R&D - Zhiyu has invested over 4.4 billion HKD in R&D from 2022 to the first half of 2025, with 74% of its employees dedicated to research and development [30][29]. - The company plans to allocate 70% of its IPO proceeds to continue R&D efforts, aiming to strengthen its technological barriers [33]. Group 5: Industry Impact - The listing of Zhiyu signifies a pivotal moment for Chinese AGI companies, representing their entry into the global pricing system as a complete commercial entity [36][37]. - This event marks a transition for Chinese large models from "following technology" to "global competition," indicating a new phase in the industry [37].

Core Insights - LMArena has successfully secured $150 million in Series A funding, raising its valuation to $1.7 billion, marking a strong start to the new year [1][3]. Group 1: Funding and Valuation - The funding round was led by Felicis and UC Investments, with participation from Andreessen Horowitz and The House Fund [3]. - The significant investment reflects the attractiveness of the AI model evaluation sector in the current market [4]. Group 2: Company Background - LMArena originated from Chatbot Arena, which was created by the open-source organization LMSYS following the emergence of ChatGPT in 2023 [5][4]. - The core team consists of highly educated individuals from top universities such as UC Berkeley, Stanford, UCSD, and CMU [6]. Group 3: Technology and Evaluation Methodology - LMArena's open-source inference engine, SGLang, has achieved performance comparable to DeepSeek's official report on 96 H100 GPUs [7]. - SGLang has been widely adopted by major companies including xAI, NVIDIA, AMD, Google Cloud, Oracle Cloud, Alibaba Cloud, Meituan, and Tencent Cloud [8]. - The primary focus of LMArena is on evaluating AI models, which they began with the launch of Chatbot Arena, a crowdsourced benchmarking platform [9][10]. Group 4: Evaluation Process - LMArena employs a unique evaluation process that includes anonymous battles, an Elo-style scoring system, and human-machine collaboration [20]. - Users input questions, and the system randomly matches two models for anonymous responses, allowing users to vote on the quality of the answers without knowing the model identities [21][22]. - The platform's Elo scoring mechanism updates model rankings based on performance, ensuring a fair and objective evaluation process [22]. Group 5: Growth and Future Plans - Since securing $100 million in seed funding, LMArena has rapidly exceeded expectations, accumulating 50 million votes across various modalities and evaluating over 400 models [25]. - The newly raised funds will be used to enhance platform operations, improve user experience, and expand the technical team to support further development [25].

Core Viewpoint - Huang Renxun emphasizes that robots are the "AI immigrants" capable of taking on jobs that humans are unwilling to do, highlighting the need for AI to support economic growth and job creation [10][11]. Group 1: AI and Robotics - Huang states that the "robot revolution" will drive economic progress and create more job opportunities while maintaining low inflation levels [11]. - He predicts that by the end of this year, robots will achieve human-level capabilities in mobility, joint movement, and fine motor skills [12]. - The development of robots requires not only visual perception but also tactile capabilities, which poses significant technical challenges [13]. Group 2: Autonomous Driving - Huang introduced the world's first open-source, large-scale autonomous driving visual-language-action (VLA) reasoning model, Alpamayo 1, and praised Tesla's FSD technology as world-class [15][16]. - NVIDIA's role is to provide a complete technology stack for companies developing autonomous vehicles, rather than manufacturing the vehicles themselves [16][20]. - The company has a high industry penetration rate, with over 1 billion vehicles on the road, and expects that millions will have strong autonomous driving capabilities in the next decade [20]. Group 3: AI Infrastructure and Memory Supply - Huang introduced NVIDIA's next-generation AI supercomputing platform, Vera Rubin, and discussed the challenges posed by rising memory prices and supply constraints [24][25]. - The company is positioned as a key player in the memory market, addressing the growing demand for high-bandwidth memory (HBM) and collaborating closely with suppliers to ensure production capacity aligns with product launches [36]. Group 4: Gaming and AI - NVIDIA upgraded its super-resolution model with the new DLSS 4.5 version, indicating a shift towards AI-driven gaming experiences [31]. - Huang predicts that future video games will be filled with AI characters, significantly enhancing realism and interactivity [32][33].

Core Viewpoint - The article highlights the advancements made by the Chinese company, Weilan Dalu, in the lawn mowing robot sector, showcasing their innovative technologies and products that are setting new industry standards, particularly in the context of the CES 2026 event [1][2][4]. Product Innovations - Weilan Dalu has introduced five new product lines at CES 2026, including the flagship X4 series designed for large, complex terrains, and the H2 series tailored for intricate yard designs [4][6]. - The company emphasizes its "Navimow standard" which includes features like "zero-turn all-wheel drive" and "no deployment, automatic mapping" capabilities, enhancing user experience and operational efficiency [6][12]. Technological Advancements - The article discusses the transition from traditional physical boundary wiring to advanced RTK positioning technology, which simplifies the deployment process for lawn mowing robots [10][11]. - Weilan Dalu's latest products feature "Network RTK" technology, allowing for accurate positioning without the need for on-site base stations, thus improving user convenience and operational reliability [24][25]. Environmental Perception - The introduction of solid-state LiDAR technology enhances the robots' ability to perceive their environment, providing detailed spatial awareness and improving obstacle detection capabilities [32][35]. - The integration of AI vision with LiDAR data allows for centimeter-level precision in identifying and navigating around obstacles, significantly improving the robots' operational effectiveness in complex yard environments [39][41]. User Experience - The new features reduce the cognitive load on users, allowing them to operate the robots with minimal setup and adjustments, thus making lawn care more accessible to a broader audience [16][17]. - The article notes that the combination of advanced technologies enables the robots to operate effectively in diverse and challenging environments, ensuring consistent performance without requiring users to adapt their yard layouts [47][54]. Industry Impact - Weilan Dalu's innovations are positioned to redefine the standards for high-end smart lawn mowing robots, addressing common issues such as lawn damage during operation and enhancing overall user satisfaction [48][51]. - The article concludes that the integration of these technologies into a cohesive framework marks a significant step towards a more reliable and efficient lawn care solution, setting a precedent for future developments in the industry [53][54].