Summary of Key Points from Conference Call Industry and Company Overview - The conference discusses advancements in the AI application industry, particularly focusing on the Cloud Code tool developed by Anthropic, which has significantly impacted programming efficiency and company valuation, now estimated between $170 billion and $180 billion [1][2][3]. Core Insights and Arguments - **Cloud Code Tool**: This tool enhances programming efficiency through context engineering, utilizing a virtual machine-like approach for context management and sandbox technology for user experience optimization. It leverages user data accumulated over three years to improve product performance [1][3][4]. - **Cost Efficiency**: AI applications, particularly through tools like Cloud Code, allow teams to complete tasks at a fraction of the traditional cost, exemplified by the ability to create a company website for just $35 in one hour [1][5]. - **AIGC Applications**: The most active area in AI-generated content (AIGC) is text processing, while image generation growth has slowed. Multimedia generation, driven by models like Google Gemini 2.5, is rapidly expanding, especially in e-commerce and live streaming [1][8][9]. - **AI App Market**: The AI app market is growing quickly but remains in its infancy, lacking a dominant app. The business model is shifting from traditional subscriptions to usage-based billing, emphasizing high-quality data over ad revenue [1][10]. - **Context Management**: Scene intelligence addresses the limitations of large models in context management, enhancing the precision of information services, such as advanced meeting record systems [1][11][12]. - **Industry-Specific AI Apps**: Despite the capabilities of large models like ChatGPT, specialized industry AI apps are necessary due to the complexity of high-quality prompt writing and context management [1][6]. - **Development Stages of AI Apps**: Most AI apps are currently at the third stage of development, indicating maturity in cloud infrastructure and context management, with some companies exploring more advanced paradigms [1][7]. Additional Important Insights - **AIGC Forms**: AIGC primarily manifests in four forms: pure text, images, multimedia (video and audio). Text applications are the most competitive, while image generation has seen a decline in demand [1][8][9]. - **User Data Utilization**: The extensive user data collected allows Cloud Code to better understand user intent, further enhancing product performance [4]. - **Market Trends**: The AI app market is characterized by a lack of leading apps, with significant potential for new entrants. The shift to usage-based pricing models reflects a broader trend in the industry [1][10]. - **Challenges in Multimedia**: The multimedia segment faces challenges such as copyright issues and model alignment, but it remains one of the fastest-growing areas [1][9]. - **AI in Document Processing**: AI tools significantly improve document processing efficiency, converting unstructured documents into structured formats, enhancing speed and accuracy [1][22]. - **Future Outlook**: The next two to three years are expected to see a rise in agent-enabled apps, similar to the mobile internet boom in the early 2010s, with substantial investment interest [1][26]. This summary encapsulates the key points discussed in the conference call, highlighting the advancements and trends in the AI application industry, particularly focusing on the impact of the Cloud Code tool and the evolving market dynamics.

Quantitative Models and Construction Methods Model: DeepSeek-R1 - **Model Construction Idea**: The DeepSeek-R1 model aims to innovate in AI technology by reducing dependency on high-end imported GPUs and enhancing cost-effectiveness and performance in global markets[5][12][30] - **Model Construction Process**: - The model is based on the DeepSeek-V3 architecture and applies reinforcement learning techniques during the post-training phase to significantly improve inference capabilities with minimal labeled data[33] - The model's performance in tasks such as mathematics, coding, and natural language inference is on par with OpenAI's o1 official version[33] - The team also introduced six distilled small models using knowledge distillation techniques, with the 32B and 70B versions surpassing OpenAI o1-mini in several capabilities[34] - The model's training cost was $5.576 million, only 1/10th of GPT-4o's training cost, and its API call cost is 1/30th of OpenAI's similar services[38] - **Formula**: $$ \text{SUE} = \frac{\text{Single Quarter Net Profit} - \text{Expected Net Profit}}{\text{Standard Deviation of Net Profit YoY Change over the Past 8 Quarters}} $$ where Expected Net Profit = Last Year's Same Quarter Actual Net Profit + Average YoY Change in Net Profit over the Past 8 Quarters[55] - **Model Evaluation**: The model is highly cost-effective and adaptable to different application environments, breaking the traditional AI industry's reliance on "stacking computing power and capital"[38][43] Model Backtesting Results - **DeepSeek-R1 Model**: - **AIME pass@1**: 9.3 - **AIME cons@64**: 13.4 - **MATH-500 pass@1**: 74.6 - **GPQA Diamond pass@1**: 49.9 - **LiveCodeBench pass@1**: 32.9 - **CodeForces rating**: 759.0[36] Quantitative Factors and Construction Methods Factor: Standardized Unexpected Earnings (SUE) - **Factor Construction Idea**: SUE is used to measure the growth potential and latest marginal changes in the prosperity of the industry and individual stocks[57] - **Factor Construction Process**: - SUE is calculated as: $$ \text{SUE} = \frac{\text{Single Quarter Net Profit} - \text{Expected Net Profit}}{\text{Standard Deviation of Net Profit YoY Change over the Past 8 Quarters}} $$ where Expected Net Profit = Last Year's Same Quarter Actual Net Profit + Average YoY Change in Net Profit over the Past 8 Quarters[55] - **Factor Evaluation**: SUE effectively measures future earnings growth and the latest marginal changes in prosperity, representing the future trend changes in the industry[57] Factor Backtesting Results - **SUE Factor**: - **2022**: -29.8% - **2023**: 15.9% - **2024**: 20.1% - **2025 YTD**: 11.0%[65]

Group 1: Technology Updates - Google has updated its Gemini 2.5 model family, introducing features like a "thinking" mechanism and multi-modal input capabilities, with a pricing standard of $0.3 per million tokens for input [2] - MiniMax launched the MiniMax-M1 model, the world's first open-source large-scale hybrid architecture inference model, supporting 1 million context inputs and 80,000 token outputs, with reinforcement learning costs reduced to $530,000 [3] - The new open-source model Kimi-Dev-72B from 月之暗面 achieved the highest score in the SWE-bench Verified programming benchmark, scoring 60.4% with only 72 billion parameters [4] Group 2: Automotive and Delivery Services - 鸿蒙智行 achieved a weekly delivery of 11,600 units, maintaining the top position among new force brands for four consecutive weeks, with flagship model 问界M8 surpassing 5,000 units in weekly deliveries [5] - 京东's CEO reported that the company has over 120,000 full-time delivery riders, with daily order volume exceeding 25 million and an average monthly income of 13,000 yuan for riders in major cities [7] Group 3: Strategic Partnerships and Collaborations - 传音控股 signed a strategic cooperation agreement with Indonesian telecom operator IOH to enhance 5G terminal penetration and improve user conversion rates [12] - 德马科技 and 上海智元新创 signed a strategic cooperation agreement to explore innovative applications of humanoid intelligent robots in logistics [11] Group 4: Financial and Market Activities - 兆芯集成's IPO application has been accepted, aiming to raise 4.169 billion yuan for new generation processor projects [16] - 曹操出行 announced its IPO plans, targeting to raise 1.853 billion HKD with a share price of 41.94 HKD [17] - 京东方A plans to acquire a 30% stake in 咸阳彩虹光电科技有限公司, with a base price of 4.849 billion yuan [18] Group 5: Industry Developments - The domestic silicon carbide IDM company 芯聚能半导体 has successfully scaled its SiC chip production for automotive applications [14] - The memory market is experiencing significant price increases for DDR4 memory, although market transactions are showing signs of weakness [15]

文 / 于钟海 , 魏鹳霏 , 肖楷 , 赵丽萍 中金研究 以MiniMax V-Triune新框架成果为例，推理感知统一框架在可拓展性、泛化性初步验证。 V-Triune以三层组件架构实现视觉推理和感知任务统一至强化学 习框架：1）多模态样本数据格式化；2）验证器奖励计算，采用异步客户端-服务器架构，奖励计算和主训练循环解耦；3）数据源级指标监控，便于溯源 和提升稳定性。结合动态IoU奖励机制、冻结ViT参数等工程优化，Orsta系列模型（32B参数）在MEGA-Bench Core基准测试中实现了最高14.1%的性能提 升。 多模态推理助力智能驾驶能力升阶。 在智能驾驶场景，多模态推理是增强道路交通标志识别判断能力、提升复杂场景泛化性的重要途径，正成为头部智 能驾驶企业算法演进的重点。2025年5月30日，蔚来世界模型NVM首个版本正式开启推送，具备全量理解、想象重构和推理能力，能够对实时环境多模信 息进行理解和推演，在选择最优ETC车道通行、停车场自主寻路等场景的性能提升显著。此外，理想自研的VLA大模型亦具备思维链推理能力，以多模态 推理模拟人类驾驶员的思维运作方式。 图表1：MiniMax多模态RL ...