Core Viewpoint - Ant Group has entered the AI healthcare sector with its product AQ, which integrates various healthcare services into a seamless experience, addressing the demand for medical consultations and related services [1][2]. Group 1: Product Features - AQ utilizes AI capabilities to create a closed-loop system for healthcare, including medical insurance, payment, and local delivery services [2][11]. - The product offers a user-friendly consultation process that mimics traditional hospital visits, providing preliminary assessments and diagnostic suggestions based on user input and image analysis [13][10]. - AQ can analyze skin conditions, heart rate abnormalities, and even traditional Chinese medicine diagnostics, showcasing its versatility [6][30][25]. Group 2: User Experience - Users report that the diagnostic results from AQ are generally accurate, often aligning with conclusions from top-tier hospitals [17][10]. - The system includes a knowledge base called AQ Intelligence, which breaks down diagnostic keywords into categories like causes, symptoms, and treatment options, enhancing user understanding [18][20]. - While the product has many strengths, some functionalities are similar to existing AI agents, raising questions about its uniqueness [11][12]. Group 3: Limitations and Concerns - Certain diagnostic results appear overly generalized, lacking personalization, which may affect user trust [22][24]. - The AI struggles with complex imaging, such as CT scans, indicating limitations in its diagnostic capabilities [36][12]. - Privacy concerns have been raised regarding the integration of personal health data within the platform [43][44]. Group 4: Overall Assessment - The integration of various healthcare functions into a single app enhances user convenience, allowing for easy appointment scheduling, medication purchases, and insurance inquiries [41][26]. - The overall user experience is reported to be smooth, with a well-structured process from diagnosis to treatment [42][40]. - Users are advised to utilize AQ for minor health issues and routine inquiries, while still recommending professional medical consultations for serious conditions [44][46].

让我们共同见证年度之星，点亮未来的方向。 组委会 发自 凹非寺 量子位｜公众号 QbitAI 为了让更多从业者感受智能浪潮的跃迁，也为了给予更多同行同路人掌声与鼓舞，我们将正式启动 「2025人工智能年度榜单」评选报名 。 这是量子位人工智能年度榜单的 第8年 。八年来，我们见证了技术的突破与落地，产业的融合与重塑，也见证了一批又一批推动时代前行的 企业、人物与产品。 在人工智能重新定义一切的时代里，智能技术已不再是单一工具，而是产业与社会协同进化的驱动力。我们期待通过这场年度评选，去发现并 致敬那些真正引领变革、开拓边界的探索者与实践者。 本次评选将从 企业 、 产品 、 人物 三大维度，设立五类奖项。欢迎企业踊跃报名！ 企业榜 产品榜 人物榜 2025 人工智能年度 焦点人物 详细评选标准及报名方式如下。 2025 人工智能年度领航企业 2025 人工智能年度 领航企业 2025 人工智能年度 潜力创业公司 2025 人工智能年度 杰出产品 2025 人工智能年度 杰出解决方案 将面向中国人工智能领域，评选出最具综合实力的企业， 参选条件 ： 评选标准 ： 2025 人工智能年度潜力创业公司 聚焦于中国人 ...

Core Viewpoint - The article highlights the rapid advancements in the AI video generation field, particularly focusing on the new features and upgrades of the Vidu platform, which aims to enhance user experience and creativity in content creation. Group 1: New Features of Vidu - The long-awaited Vidu Q2 reference generation feature is officially launched, allowing for high consistency, faster processing, and more affordable pricing without the need for an invitation code [2][13]. - Vidu's video extension feature allows users to extend videos up to five minutes, with free users able to generate videos up to 30 seconds [20]. - The Vidu app has undergone a comprehensive redesign, transforming from an AI creation platform to a one-stop AI content social platform, enabling users to easily create and share videos [4][12]. Group 2: User Experience Enhancements - Users can create engaging duet videos by simply tagging a subject and providing a brief prompt, significantly lowering the creative barrier [7]. - The app includes a vast library of subjects, including characters and effects, allowing users to generate fun videos anytime and anywhere [8]. - The platform now supports browsing various AI-generated video content, enhancing the social aspect of video sharing [9]. Group 3: Performance Improvements - Vidu Q2 shows a threefold increase in generation speed compared to the previous version, allowing creators to transform ideas into videos more efficiently [40]. - The platform maintains high video quality, ensuring that even demanding scenarios like animation and advertising are well-handled [25]. - The combination of high consistency, video extension capabilities, and 1080P resolution meets the needs of content creators and companies for quality AI video generation [24]. Group 4: Commercial Applications - The advancements in Vidu's technology significantly lower the production costs and barriers for marketing videos, making it accessible for small and medium-sized businesses [47]. - A typical application scenario in the e-commerce sector allows merchants to create dynamic product showcase videos quickly by providing static images and simple prompts [43][46]. - The democratization of technology is expected to unleash creativity among users, enabling anyone to generate high-quality videos with minimal effort [47].

Core Insights - The article introduces Mem-α, an innovative reinforcement learning framework designed to enable large language models (LLMs) to autonomously manage complex memory systems, moving away from reliance on manual design and predefined instructions [2][4][14]. Memory Management Challenges - Traditional memory-enhanced agents often depend on predefined instructions and tools for memory updates, which can lead to suboptimal memory construction and information loss, particularly in long-term interactions [7][9][8]. - LLMs face limitations due to finite context windows, making external memory systems crucial for understanding long-term information [5][6]. Mem-α Framework - Mem-α transforms the memory construction problem into a sequential decision-making problem that can be optimized through reinforcement learning, allowing agents to explore optimal memory management strategies during information processing [14][16]. - The framework incorporates a complex memory system inspired by cognitive science, consisting of core memory, episodic memory, and semantic memory, each supporting various memory operations [22][20]. Training and Evaluation - Mem-α utilizes a multi-dimensional reward function to optimize memory construction, focusing on accurate retrieval, test-time learning, long-range understanding, and conflict resolution [18][28]. - Experimental results demonstrate that Mem-α significantly outperforms existing methods, achieving higher accuracy and efficient memory usage while maintaining performance [35][36]. Key Findings - Mem-α shows superior performance across all tasks, particularly in accurate retrieval and long-range understanding, indicating strong generalization capabilities [35]. - The framework reduces memory usage by approximately 50% compared to traditional methods while enhancing performance, validating the effectiveness of semantic compression mechanisms [35]. - The structured architecture of Mem-α proves essential for processing complex information, highlighting the limitations of flat memory representations [35]. - Mem-α exhibits robust generalization to document lengths exceeding 400K tokens, despite being trained on documents averaging less than 30K tokens [35].

衡宇 发自 凹非寺 量子位 | 公众号 QbitAI 宇树第四款人形机器人， Unitree H2 转着圈圈来了！ 这款新品身高180厘米，体重70公斤，比同身高的H1机器人重了足足23公斤。 宽肩窄腰，意味着电池和控制板都必须塞进它小小的胸膛。 相较于前作， 宇树H2 的最大变化是在 外观上增加了仿生人脸 。 从长相到身高体重，H2整体形态更接近真人。 不过，广大网友们对这张脸的美感好像不太买账…… 虽然看起来神似2004年威尔·史密斯主演科幻电影《我，机器人》 （又译为《机械公敌》） 里的机器人NS-5。 但刷遍各大平台评论区， 大家都觉得它有点诡异 。 不知道是不是美瞳直径太大，惹得大家恐怖谷效应犯了。 | (a^ X )ndrew � 2 @0xnullcline · 43m | | --- | | fire whoever designed the face please | | Tru (") - 111 470 | 截至发稿，宇树官网上还没有更新H2的详细信息。 只有官号放出宣传片时有一小段配套文案介绍： （H2定位为） 仿生人形机器人，为每个人安全友好地服务而生。 | 型号 | HI | G1 ...

Core Viewpoint - OpenAI is facing a significant scarcity of computing power, which is critical for innovation in the AI field [1][2][4] Resource Allocation Mechanism - OpenAI has a structured yet challenging resource allocation mechanism for its limited computing resources [8] - Resources are divided between research and application sides, with major decisions made by the executive team [9][10] - Within the research domain, allocation is determined by the chief scientist and research director [12] - A team led by Kevin Park manages the reallocation of idle GPUs to meet the demands of various projects [14][15] Industry Implications - The internal competition for computing resources at OpenAI reflects the broader dynamics of the AI industry, where computing power directly influences AI capabilities [16][17] - The founder of AI chip company Groq emphasized that controlling computing power equates to controlling AI [18] - OpenAI's computing power expenditure reached $7 billion last year, and the company is now building its own data centers, achieving nearly a trillion in computing transactions [19][20] Competitive Landscape - The competition for computing resources is not only internal but also extends to the entire AI computing market [20] - Meta's CEO, Mark Zuckerberg, highlighted the importance of computing resources as a competitive advantage for researchers [22] - The future of AI development places computing power at the forefront of strategic importance [23]

Core Insights - The article discusses the evolution of OpenAI's models, particularly focusing on GPT-5 as an iteration of the o3 model, suggesting that it represents a significant advancement in AI capabilities [1][4][23]. Model Evolution - Jerry Tworek, OpenAI's VP of Research, views GPT-5 as an iteration of o3, emphasizing the need for a model that can think longer and interact autonomously with multiple systems [4][23]. - The transition from o1 to o3 marked a structural change in AI development, with o3 being the first truly useful model capable of utilizing tools and contextual information effectively [19][20]. Reasoning Process - The reasoning process of models like GPT-5 is likened to human thought, involving calculations, information retrieval, and self-learning [11]. - The concept of "thinking chains" has become prominent since the release of the o1 model, allowing models to articulate their reasoning in human language [12]. - Longer reasoning times generally yield better results, but user feedback indicates a preference for quicker responses, leading OpenAI to offer models with varying reasoning times [13][14]. Internal Structure and Research - OpenAI's internal structure combines top-down and bottom-up approaches, focusing on a few core projects while allowing researchers freedom within those projects [31][33]. - The company has rapidly advanced from o1 to GPT-5 in just one year due to its efficient operational structure and talented workforce [33]. Reinforcement Learning (RL) - Reinforcement learning is crucial for OpenAI's models, combining pre-training with RL to create effective AI systems [36][57]. - Jerry explains RL as a method of training models through rewards and penalties, similar to training a dog [37][38]. - The introduction of Deep RL by DeepMind has significantly advanced the field, leading to the development of meaningful intelligent agents [39]. Future Directions - Jerry believes that the future of AI lies in developing agents capable of independent thought for complex tasks, with a focus on aligning model behavior with human values [53][54]. - The path to AGI (Artificial General Intelligence) will require both pre-training and RL, with the addition of new components over time [56][58].

Core Viewpoint - The article discusses the emergence of a new AI assistant application named Cici, developed by ByteDance, which has rapidly gained popularity in various countries, indicating a competitive landscape in the AI assistant market. Group 1: Cici's Rise and Features - Cici has achieved significant download growth, ranking as the top downloaded app in Mexico's Google Play Store and within the top 10 free apps in the UK Apple App Store [2] - The application utilizes technologies from ByteDance's other platforms, including image editing and code assistance tools, and incorporates OpenAI's GPT models and Google's Gemini for chat generation [8][9] - Cici's interface design is similar to that of Doubao, another ByteDance product, and it allows users to interact via text or voice, supporting image generation and analysis [10] Group 2: Competitive Landscape in AI Assistants - Doubao has maintained a dominant position in the domestic AI assistant market, with a cumulative download exceeding 100 million, while other competitors like Kimi, DeepSeek, and Tencent Yuanbao follow behind [16][22] - The top four AI assistant products, including Doubao, account for approximately 93% of the user base in the market, showcasing a significant "Matthew Effect" [17][24] - In terms of daily active users (DAU), Doubao leads with 33 million, followed by DeepSeek and Tencent Yuanbao with 25 million and 16 million respectively [23] Group 3: ByteDance's Global Strategy - The success of Cici reflects ByteDance's strategy to expand its AI capabilities globally, with a focus on specific markets such as the UK, Mexico, and Southeast Asia [12] - Despite Doubao's comprehensive lead in various dimensions, DeepSeek remains strong in the web-based AI assistant segment, indicating a competitive challenge for ByteDance [27]

该框架在4B及以下的Qwen、Gemma上已被证实有效，理论上可用于其他Transformer模型。 克雷西 发自 凹非寺 量子位 | 公众号 QbitAI 1.58bit量化，内存仅需1/10，但表现不输FP16？ 微软最新推出的蒸馏框架 BitNet Distillation （简称BitDistill），实现了几乎无性能损失的模型量化。 同等硬件性能下，使用该方法量化后的 推理速度提升2.65倍，内存消耗仅1/10 。 网友看了之后表示，如此一来昂贵的GPU将不再是必需品，英伟达的好日子要到头了。 BitDistill框架设计 BitDistill包含三个依次衔接的阶段，分别是 模型结构优化 （Modeling Refinement）、 继续预训练 （Continue Pre-training）和 蒸馏式微 调 （Distillation-based Fine-tuning）。 建模结构优化的主要目标是为1.58-bit模型训练提供结构层面的支持，缓解低精度训练中常见的优化不稳定问题。 在传统的全精度Transformer模型中，隐藏状态的方差通常在预训练时已被良好控制。然而，当模型被压缩到极低位宽（如 ...

美团LongCat团队投稿 发自 凹非寺 量子位 | 公众号 QbitAI 美团LongCat团队发布了当前高度贴近真实生活场景、面向复杂问题的大模型智能体评测基准—— VitaBench （Versatile Interactive Tasks Benchmark）。 VitaBench以 外卖点餐、餐厅就餐、旅游出行 三大高频生活场景为典型载体，构建了一个包含 66个工具 的交互式评测环境，并设计了跨场 景综合任务。 例如，在旅行规划任务中，要求智能体通过推理、调用工具与用户交互，完整完成从购票到预订餐厅的全流程。 团队首次从深度推理、工具使用与用户交互三大维度对智能体任务进行量化拆解，从而实现对复杂问题的可控构建。 评测结果显示，即便是当前先进的推理模型，在主榜（复杂跨场景任务）上的成功率也仅约 30% ，揭示了现有智能体与真实生活应用需求之 间的显著差距。 目前，VitaBench已全面开源，旨在为推动智能体在真实生活场景中的研发与落地提供重要基础设施。 研究背景：智能体评测与现实应用间存在巨大鸿沟 随着大语言模型在复杂推理与工具调用能力上的快速进步，基于LLM的智能体在真实生活场景中的应用日益广泛。 ...