Core Viewpoint - The article discusses the contrasting approaches of China and Japan in the robotics industry, highlighting China's rapid advancements and Japan's historical challenges in humanoid robotics, while emphasizing Japan's strength in industrial robotics [5][7][17]. Group 1: Industry Overview - Robotics has become a significant focus for high-tech startups in China since 2020, alongside AI, with substantial investments and policy support [5]. - Japan, despite being an early innovator in humanoid robotics, has struggled to commercialize its developments, with notable failures like ASIMO and Pepper, which faced high costs and low market demand [9][10]. - The global market for industrial robots is projected to account for approximately 71.4% of total revenue in 2024, indicating a strong demand for industrial applications [21]. Group 2: Comparative Analysis - Japan's robotics industry has historically focused on stability and reliability, with a strong emphasis on long-term partnerships and high-quality components, while China's approach is characterized by rapid innovation and cost competitiveness [29][32]. - The article notes that Japan has lost its exploratory drive in humanoid robotics, while China is making significant strides in this area, leveraging advancements in AI and manufacturing capabilities [10][17]. - The differences in robotics strategies between China and Japan are attributed to varying economic structures, technological foundations, and risk preferences [19][32]. Group 3: Market Dynamics - The article highlights that the demand for humanoid robots is not yet strong globally, with most buyers being experimental or marketing-oriented, which complicates the commercial viability of humanoid robotics [16][19]. - Japan's industrial robotics sector remains robust, with major companies like FANUC and Yaskawa maintaining significant market shares and focusing on B2B business models [22][24]. - The article emphasizes that Japan's industrial robots are integrated with advanced AI technologies, enhancing their operational capabilities while maintaining a focus on reliability and precision [27][30].

今年，华泰证券与南京大学开展合作，以多元方式鼓励青年科技人才成长。公司一方面在南京大学设立 华泰证券科技奖学金，奖励计算机、人工智能、数学等专业的优秀学生；另一方面与人工智能、新闻传 播等学院深度合作，在AI赋能金融、AI创新内容生成和传播等多领域开展前瞻性探索，结合行业前沿 视角，为青年人创意的生长、展示和落地提供平台。 本次AI+创新创意大赛是双方首次联合举办人工智能领域的学生竞赛。自6月启动以来，针对华泰证券 提出的大模型上市公司财报分析、金融场景 API 工具调用研究、金融视觉大模型应用研究这三大主 题，学生们给出了多样化的解决方案。南京大学人工智能学院和华泰证券信息技术部专家还共同进行了 前期评审和辅导培训，帮助学生们进一步完善方案。最终，"慧视解图：金融文档图表智能问答系统"等 六个团队带来的解决方案分获一、二、三等奖。赛后，学生们还参观了"数字华泰"运营指挥中心，对科 技如何应用于实际金融场景有了更为具象的感受。 他们基于人工智能及大模型技术，尝试用创新方式解决金融业务场景中的真实问题。作为行业数字化转 型领跑者，华泰证券始终秉持守正创新的精神，不断弘扬践行中国特色金融文化，通过不断加强数字化 ...

Core Insights - The article emphasizes the establishment of a comprehensive community for autonomous driving knowledge, aiming to facilitate learning and collaboration among industry professionals and newcomers [22][10][8] Group 1: Community and Learning Resources - The "Autonomous Driving Heart Knowledge Planet" has over 4,000 members and aims to grow to nearly 10,000 in two years, providing a platform for technical exchange and job opportunities [8][22] - The community offers a variety of resources, including video tutorials, learning routes, and Q&A sessions, to help members navigate the complexities of autonomous driving technology [10][23] - Members can access insights from industry leaders and academic experts, enhancing their understanding of the latest trends and technologies in autonomous driving [11][10] Group 2: Technical Insights and Developments - Recent updates include discussions on Waymo's latest base model, advancements in self-driving technology, and insights from industry conferences [7][10] - The community has compiled over 40 technical routes covering various aspects of autonomous driving, such as perception, simulation, and planning control [23][10] - Key topics include end-to-end autonomous driving, multi-modal large models, and the integration of traditional planning with new technologies [44][52][56] Group 3: Job Opportunities and Industry Trends - The community provides job recommendations and internal referrals to help members connect with leading companies in the autonomous driving sector [27][10] - Members can inquire about job openings, industry trends, and the future of autonomous driving technologies, fostering a supportive environment for career development [26][10] - The platform encourages collaboration between academia and industry, aiming to bridge the gap between research and practical applications in autonomous driving [22][11]

编辑｜+0、泽南、Panda LLM 智能体很赞，正在成为一种解决复杂难题的强大范式。 论文标题：Detailed balance in large language model-driven agents 论文地址：https://arxiv.org/pdf/2512.10047 简单来说，他们通过实验测量了 LLM 生成状态之间的转移概率。基于此，他们在统计上发现了 LLM 生成转移中的细致平衡 (detailed balance) 现象。 这表明： LLM 的生成可能不是通过一般性地学习规则集和策略来实现的，而是通过隐式地学习一类潜在的势函数 (potential functions)，这些势函数可能超越了不 同的 LLM 架构和提示词模板。 不过，这种成功目前更多还停留在「经验主义」的工程实践层面 —— 我们知道它好用，但往往不知道它在宏观上为何如此运作。那么，我们是否能找到一个理论 框架，像物理学描述自然界那样，去理解和统一智能体的宏观动力学（macroscopic dynamics）？ 为了解开这个黑盒，近日，北京大学物理学院、高能物理研究中心以及北京计算科学研究中心联合发力，跨界借用了物理学中经 ...

Core Insights - The future of AI architecture is expected to surpass the capabilities of the current Transformer model, potentially enhancing intelligence by a factor of 10,000 [72]. Group 1: Conference Overview - The 8th GAIR Global Artificial Intelligence and Robotics Conference commenced in Shenzhen, focusing on the intersection of academia, industry, and investment in AI [3]. - The conference serves as a platform for high-quality discussions and insights into the forefront of AI technology, reflecting on the rapid transformation driven by large models over the past four years [3]. Group 2: Key Technological Developments - The LongLoRA technology was introduced in 2023, marking the world's first 32K long-text context understanding model [5][13]. - The Mini-Gemini platform, launched in 2024, gained over 3,000 stars on GitHub and is recognized as the strongest model in the open-source community, integrating multimodal understanding capabilities [5][18]. - A new version of Mini-Gemini was released, featuring a complete Chinese voice system capable of long video comprehension and cross-language generation [5][20]. Group 3: Innovations in Image Generation - The ControlNeXt technology allows lightweight operations for image style transfer and dynamic effect generation [6]. - The DreamOmni2 system, developed with significantly fewer resources than competitors, is positioned as a leading unified system for intelligent image generation and editing [6][36]. - DreamOmni2 can perform complex tasks such as virtual try-ons, image editing, and product design, demonstrating capabilities that may surpass existing tools like Photoshop [37][40]. Group 4: Future Directions in AI - The development of large models should focus on improving the connectivity of neurons rather than merely increasing their quantity, emphasizing the importance of neural connections and brain complexity [7][70]. - Future AI training methods are expected to shift from one-time learning to continuous learning, akin to human education, which will enhance the adaptability and intelligence of AI systems [75]. - The integration of robotics and physical embodiments into AI systems is seen as crucial for bridging the gap between AI and human-like understanding [75].

Core Insights - The launch of Inspur Technology's Smart Base Platform V4.0 focuses on five key areas: integration of large and small model technologies, enhancement of trusted data circulation, promotion of integrated operational management, improvement of intelligent research and operation levels, and creation of a unified industry knowledge base [1] Group 1: Data Security and Trustworthiness - The platform has established a comprehensive data circulation technology support system covering the entire lifecycle, ensuring security and reliability at every stage from sharing to trading [3] - It integrates technologies for trusted identity, trusted environment, secret storage, trusted transmission, and secure computing, providing full privacy protection for data during static storage, dynamic transmission, and collaborative computing [3] - The platform's data resource governance and intelligent data entities cover the entire process from collection and quality governance to sharing applications, ensuring data resources are visible, traceable, manageable, controllable, and usable [3] Group 2: High-Quality Data Support for Large Models - The platform features a multi-agent collaborative operating system equipped with a self-planning intelligent agent engine, capable of automatically breaking down complex tasks and dynamically optimizing execution strategies [4] - It deeply integrates multi-modal analysis technology to structure and cross-analyze text, visual, and voice data across various industries such as construction, public security, taxation, emergency response, and forestry [4] - The self-developed high-quality corpus governance platform V1.0 establishes a proactive multi-modal corpus governance system, enabling continuous production of targeted optimized corpus [4] Group 3: Digital Operation Maturity - The platform has upgraded its operational functionality to achieve three main goals: standardized security, closed-loop management, and intelligent decision-making [5] - It has identified over 50 standardized operational matters and established a comprehensive evaluation index system with more than 100 indicators, providing quantifiable, traceable, and pre-warning intelligent operational management support [5] - The platform offers an integrated knowledge base solution that connects the entire process of knowledge production, aggregation, management, and application, enhancing knowledge management towards a structured and intelligent direction [5]

Core Insights - Alibaba has launched the new Tongyi Wanshang 2.6 series model, which includes five new models that enhance capabilities in video and image generation, covering various creative processes from single-use generation to reusable creation [2][5] - The Wanshang 2.6 model is the first in China to support character role-playing in video generation, with improvements in video quality, sound effects, and adherence to instructions, achieving a maximum video length of 15 seconds [2][4] Model Features - The Wanshang 2.6 model integrates multiple innovative technologies for multi-modal joint modeling and learning, allowing it to extract and maintain consistency across visual and auditory features during video generation [7][9] - It can convert simple user prompts into multi-scene scripts, generating coherent narrative videos while maintaining consistency in key elements like subjects and scenes [9][11] User Experience - Users can upload personal videos and input prompts to quickly generate narrative videos with cinematic quality, enabling anyone to take on a director's role [9][11] - The model supports various applications, including AI comic creation, advertising design, and short video production, with over ten visual creation capabilities available [12] Image Generation Enhancements - The model has improved in style control and expression stability, allowing for better integration and transition between different artistic styles while reducing the "AI feel" in generated realistic portraits [13][15] - It can generate posters, illustrations, or infographics based on longer, structured text, enhancing the clarity of the relationship between content and visuals [15][19]

Core Viewpoint - The Hong Kong stock market has weakened due to various factors, but there is a counter-trend investment in the technology sector as it transitions from emotional valuation recovery to value creation, indicating potential for future profit growth [1][9]. Group 1: Market Dynamics - Recent adjustments in the Hong Kong stock market are influenced by the return of southbound funds due to new public offering regulations, the peak of IPO unlocks, and strong expectations of interest rate hikes by the Bank of Japan affecting overseas liquidity [1][9]. - Since November 2025, ETFs tracking the Hang Seng Tech Index have seen a total inflow of 27 billion yuan, with the Hang Seng Tech ETF (513130) alone attracting 4.7 billion yuan over 25 trading days, reflecting a significant increase in investor interest [10]. Group 2: Financial Performance - The Hang Seng Tech Index's latest price-to-earnings (P/E) ratio stands at 23.13, which is at a mid-low percentile of 34.20% over the past five years, suggesting that the index may offer attractive valuation compared to major tech indices in A-shares and US markets [12]. - Internet giants have reported a general recovery in revenue, with "AI + Cloud" businesses highlighted as key growth areas, indicating that AI strategies are beginning to contribute to revenue [10]. Group 3: Policy Support - The Central Economic Work Conference has prioritized "innovation-driven development" as a key task for the upcoming year, focusing on building international innovation centers and enhancing "AI +" and industrial upgrades, providing solid policy support for the tech sector [11]. Group 4: Future Outlook - With the anticipated easing of monetary policy by the Federal Reserve and improving overseas liquidity, combined with the stable performance of domestic tech leaders since Q3 2025, the Hong Kong tech sector is expected to enhance its medium to long-term investment value [13]. Group 5: Investment Tools - The Hang Seng Tech ETF (513130) closely tracks the Hang Seng Tech Index, which includes 30 leading Hong Kong internet and tech companies across various sectors, making it a comprehensive and representative investment tool for the AI industry chain [15]. - The ETF offers advantages such as large scale, good liquidity, and support for T+0 trading, with a low management fee of 0.2% per year, making it an important tool for investors looking to access core tech assets in Hong Kong [16].

Core Viewpoint - The Hong Kong stock market is experiencing a downturn, but the technology sector is seeing a counter-trend investment as the AI sector moves towards value creation and commercialization, indicating potential for future profit growth [1] Group 1: Market Trends - Despite the recent weakness in the Hong Kong stock market, there is a notable influx of funds into the technology sector, reflecting market recognition of the long-term fundamentals [1] - Since November 2025, ETFs tracking the Hang Seng Technology Index have seen a total of 27 billion yuan in net inflows, with the Hang Seng Technology ETF (513130) attracting 4.7 billion yuan and increasing its shares by 6.2 billion [1] - The Hang Seng Technology Index's current price-to-earnings (P/E) ratio is 23.13, which is at a mid-low percentile level compared to the past five years [2] Group 2: Comparative Valuation - The valuation of the Hang Seng Technology Index is lower than that of major technology indices in the A-share and US markets, such as the Nasdaq Index at 41.09 and the Sci-Tech Innovation 50 Index at 156.57 [2] - The recent market correction has released some risk factors, making the technology sector's valuation more attractive for investment [2] Group 3: Future Outlook - With the anticipated interest rate cuts by the Federal Reserve and improving overseas liquidity, along with stable performance from domestic tech leaders since Q3 2025, the long-term investment value of the Hong Kong technology sector is expected to increase [2] - The Hang Seng Technology ETF (513130) closely tracks the Hang Seng Technology Index, which includes 30 leading Hong Kong internet and tech companies across various sectors, providing comprehensive exposure to the AI industry chain [2] Group 4: Investment Tools - The Hang Seng Technology ETF (513130) offers advantages such as large scale, good liquidity, and support for T+0 trading, making it a key tool for investors looking to access core assets in the Hong Kong technology sector [3] - Investors can also consider the off-market linked funds (Class A 015310, Class C 015311) for additional investment opportunities [3]

宗建树 陈耀文 SAC：S0490520030004 SAC：S0490525070002 SFC：BUX668 $${\mathrm{iii}}\{8\}{\mathrm{iii}}\{8\}{\mathrm{iii}}\{8\}{\mathrm{iii}}\{8\}{\mathrm{iii}}\{8\}{\mathrm{iii}}\{8\}$$ 丨证券研究报告丨 行业研究丨点评报告丨软件与服务 [Table_Title] AI 产业跟踪：GPT-5.2 发布，关注大模型迭代进 展 报告要点 [Table_Summary] 当地时间 12 月 11 日，OpenAI 正式发布新一代大模型 GPT-5.2，主打通用智能、编码能力和 长上下文任务处理，作为近期连续版本更新中的又一项核心升级。此次，OpenAI 发布"迄今为 止功能最强大的专业知识工作模型系列"——GPT-5.2，在提升性能的同时，将重心逐步转移 至其模型货币化，或将加速大模型的商业化落地。建议关注：1）国内大模型厂商；2）国内大 型云厂商；3）垂类场景 Agent 厂商；4）国产算力产业链。 分析师及联系人 [Table_Author] %% ...