Core Viewpoint - Teslin, established in 2015, is a key player in China's AIoT industry, focusing on technology-driven industrial upgrades and spatial intelligence for sustainable development [1] Financial Performance - Teslin's revenue for 2024 is projected to be 1.843 billion yuan, representing an 83.2% increase compared to 2023 [1][2] - Revenue figures for 2022 and 2023 were 738 million yuan and 1.006 billion yuan, respectively, resulting in a compound annual growth rate (CAGR) of 58.0% from 2022 to 2024 [1][2] - The company's expense ratio (sales, management, and R&D) decreased from 76.9% in 2023 to 45.0% in 2024, indicating effective cost control [3] Market Position - Teslin has become one of the fastest-growing companies in the AI industry, outperforming peers such as SenseTime and Horizon Robotics, which reported revenue growth rates of 10.8% to 53.6% in 2024 [3] - The company has established a comprehensive AIoT technology product system over nine years, positioning itself as a leading enterprise in the rapidly growing AIoT market [2] Market Expansion - As of December 31, 2024, Teslin's products have been deployed by over 800 clients across 160 cities globally, with a total order amount of 2.3 billion yuan [4] - The number of clients increased from 224 in 2022 to 342 in 2024, reflecting an optimized customer structure [4] Strategic Focus - Teslin is focusing on three strategic directions: AIoT models, AIoT infrastructure, and AIoT agents, which are expected to drive future business growth [6] - The company is responding to the increasing demand in the market by restructuring its internal teams to enhance efficiency and innovation [3] Industry Context - The global AIoT market is experiencing rapid growth, with a projected CAGR of over 31.7% over the next five years [2] - China's AI market is also expanding, with spending reaching 14.8 billion USD in 2023, making it the second-largest AI market globally [7] - Teslin's technology strategy aligns with China's push for self-sufficiency in AI, reducing reliance on external technologies and enhancing the resilience of the industrial chain [7]

DeepSeek放大招！新模型专注数学定理证明，大幅刷新多项高难基准测试。 在普特南测试上， 新模型 DeepSeek-Prover-V2 直接把记录刷新到 49道 。 目前的 第一名 在657道题中只做出 10道 题，为Kimi与 AIME2024冠军团队Numina 合作成果 Kimina-Prover 。 而未针对定理证明优化的 DeepSeek-R1只做出 1道 。 让还没发布的R2更令人期待了。 | 657) | | --- | | (out of | | Lean | | मै | Model | num- | | | --- | --- | --- | --- | | | | solved | compute | | 1 | Kimina-Prover-7B-Distill♥ | 10 | pass@192 | | 2 | Self-play Theorem Prover♥ | 8 | pass@3200 | | 3 | Goedel-Prover-SFT♥ | 7 | pass@512 | | 4 | ABEL | 7 | pass@596 | | 5 | InternLM2.5-StepPr ...

Core Insights - DeepSeek has released DeepSeek-Prover-V2, an open-source large language model specifically designed for formal theorem proving, achieving industry-leading performance in theorem proving tasks [1][3][4]. Model Overview - Two versions of DeepSeek-Prover-V2 have been released, with parameter sizes of 7 billion and 671 billion. The larger model is based on DeepSeek-V3-Base, while the smaller one is built on DeepSeek-Prover-V1.5-Base, supporting a maximum context length of 32,000 tokens [3][4]. - DeepSeek-Prover-V2 is tailored for the mathematical AI programming language Lean 4, focusing on formal theorem proving [3][4]. Technical Implementation - The model utilizes a recursive theorem proving process to generate cold-start training data, where DeepSeek-V3 decomposes complex problems into manageable sub-goals and formalizes the reasoning steps [9][11]. - The training process involves two phases: a non-CoT (non-Chain of Thought) mode for rapid formal proof generation and a CoT mode for detailed reasoning steps, enhancing transparency and logical progression [17][19]. Performance Metrics - The DeepSeek-Prover-V2-671B model achieved an 88.9% pass rate on the MiniF2F test and successfully solved 49 out of 658 problems in the PutnamBench dataset [15][23]. - The model's performance was evaluated against various benchmarks, demonstrating unprecedented accuracy and efficiency compared to other advanced models in the industry [20][23]. Dataset Release - DeepSeek has also introduced ProverBench, a benchmark dataset containing 325 problems, including 15 from recent AIME math competitions, aimed at comprehensive evaluation of models in high school and undergraduate mathematics [25][26].

Core Viewpoint - DeepSeek has launched DeepSeek-Prover-V2-671B, marking a significant advancement in AI mathematical reasoning capabilities, particularly in automated theorem proving [2][4]. Group 1: Model Overview - DeepSeek-Prover-V2-671B is a next-generation automated theorem proving expert model with 671 billion parameters, optimized for proof generation and verification in the Lean 4 framework [4][6]. - The model employs a mixture of experts (MoE) architecture, activating approximately 37 billion parameters per inference, enhancing computational efficiency while maintaining strong reasoning capabilities [4][6]. Group 2: Key Breakthroughs - The release signifies three major milestones, including the potential for innovation across various application domains [6]. - The model's specifications include a context length of approximately 128,000 tokens, allowing it to handle complex reasoning chains and lengthy proofs [6][7]. - The attention mechanism is likely a multi-head latent attention (MLA), which compresses key-value (KV) cache, significantly reducing memory requirements [6][7]. Group 3: Applications and Impact - The model supports formal verification in areas such as cryptographic security proofs and chip design validation, enabling rigorous mathematical checks in automated processes [7]. - It aids mathematicians in formalizing theorems, exploring new conjectures, and proving complex mathematical problems, potentially accelerating mathematical research [7]. - The model can be utilized as an interactive educational tool, guiding students in mastering rigorous mathematical proof methods [7].

Group 1 - The U.S. economy showed unexpected contraction with a GDP decline of 0.3% in Q1, marking the first quarterly negative growth since 2022, significantly below the expected growth of 0.4% [3][11] - The market reacted sharply to the GDP data, with the Dow Jones dropping nearly 800 points and the Nasdaq falling close to 3% during early trading [3][5] - Following news of potential trade negotiations and tariff adjustments, the market began to recover, reducing most of the earlier losses [5] Group 2 - Major companies like AMD saw significant stock declines, with Supermicro Computer dropping over 11% due to disappointing earnings forecasts [6] - Among the tech giants, Microsoft and Meta reported better-than-expected earnings, with Meta raising its capital expenditure guidance for the year, leading to stock price increases [8] - In contrast, Tesla and Amazon experienced stock declines, with Tesla down 3.4% and Amazon down 1.6% after initial larger drops [8] Group 3 - The U.S. Commerce Department indicated that the GDP decline was primarily due to a 36% surge in preemptive imports before the implementation of Trump's tariff policies, which expanded the trade deficit [11] - Consumer spending growth was weak at only 1.8%, the lowest since mid-2023, contributing to the economic slowdown [11] - Some economists warned that if current tariff policies remain unchanged, the U.S. economy could face stagnation, with a 90% probability of recession predicted by Apollo Global Management's chief economist [12] Group 4 - Trump quickly attributed the economic downturn to his predecessor, claiming it was "Biden's stock market" and asserting that the economy is merely in a transitional phase [14][15] - Criticism arose regarding Trump's economic policies, with some analysts noting that the stock market and dollar performance during his term has been the worst since 1980 [16] - Retail giants like Walmart and Target have begun to absorb tariff costs, indicating a shift in strategy due to supply chain disruptions caused by the tariff war [19][20]

Group 1 - Apple is restructuring its global affairs and music departments, including management adjustments in Europe, India, China, and other Asian regions [3] - OpenAI has rolled back the latest update of GPT-4o due to concerns about its overly flattering personality, with plans for further improvements [4] - Meituan announced that starting May 1, 2025, it will waive delivery cabinet fees for its riders, enhancing their delivery rights [8] Group 2 - Starbucks plans to increase its workforce and reduce investment in automation, aiming to improve customer experience [10] - Volvo is initiating a cost-cutting plan totaling 18 billion Swedish Krona, which includes global layoffs to enhance profitability [11] - Microsoft CEO stated that 30% of the company's code is now generated by AI, indicating a growing reliance on artificial intelligence in software development [11] Group 3 - Extreme Stone Automotive denied rumors about its founder's alleged departure, affirming that operations are normal and the founder is fulfilling his duties [13] - Decathlon is reportedly looking to sell about 30% of its Chinese business, with a potential valuation of around $1 billion [13] - Nvidia's CEO emphasized the need for AI factories in all American companies, which will create technology jobs [19] Group 4 - CATL is planning to initiate its Hong Kong listing next month, potentially becoming the largest stock issuance in the city in four years, aiming to raise at least $5 billion [20] - Great Wall Heavy Industry has completed a 520 million yuan Series A financing round, led by two Fortune Global 500 companies [20] - DeepSeek released a new AI model with 671 billion parameters, enhancing its capabilities for complex mathematical proofs [25]

Core Insights - DeepSeek has officially released DeepSeek-Prover-V2 on Hugging Face, continuing its open-source momentum with two versions launched [1][4] - The training core of DeepSeek-Prover-V2 combines "recursion + reinforcement learning," enabling the model to break down complex theorems into sub-goals and reasoning paths [3][8] Model Specifications - DeepSeek-Prover-V2-7B is based on the previous V1.5 model and supports a maximum context input of 32K [4] - DeepSeek-Prover-V2-671B is trained on the DeepSeek-V3-Base, showcasing the strongest reasoning performance [4] Training Process - The training process consists of two phases: the first phase focuses on rapid mode using an "expert iteration" method, where successful answers refine the model [5] - In the second phase, more complex logical reasoning capabilities are trained, incorporating mathematical knowledge from DeepSeek-V3 and formal data [6] Reinforcement Learning - The GRPO reinforcement learning algorithm is introduced to enhance reasoning capabilities, allowing the model to autonomously learn to select optimal solutions from multiple candidates [8] - The system generates 32 different proof schemes for each theorem, retaining only those verified as correct by the Lean verification system [9] Model Distillation - After developing the powerful 671B model, the team distilled its capabilities into a smaller 7B model, allowing users to achieve near-equivalent mathematical reasoning abilities on resource-limited devices [10][11] Reasoning Modes - The rapid mode (non-CoT) focuses on speed, generating concise Lean code answers without showing the thought process, suitable for handling numerous problems [12] - The logical mode (CoT) details each step of the reasoning process, ensuring clarity and transparency [12] Performance Evaluation - In the final performance assessment, DeepSeek-Prover-V2-671B achieved an 88.9% pass rate in the MiniF2F test, successfully solving 49 problems from the PutnamBench dataset [17] New Dataset - DeepSeek introduced a new formal mathematical dataset, ProverBench, containing 325 problems across various mathematical domains, including number theory, algebra, and calculus [18][19] Comparison and Trends - The comparison shows a significant trend: the performance gap between large language models in "informal mathematical reasoning" and "formal mathematical reasoning" is narrowing [21] - The evolution of model structure and training strategies enables models to produce rigorous, verifiable mathematical proofs [22] Future Directions - DeepSeek-Prover-V2 indicates a shift in focus from merely generating content to generating structured logic, which may touch upon the foundational structure of general artificial intelligence [33][34]

中国电子 China (Overseas) Technology 国产开源模型千帆竞发，阿里 Qwen-3、小米 MiMo、DeepSeek Prover 集中发布 A surge of domestic open-source models 姚书桥 Barney Yao 吴叡霖 Louis Ng wo[Table_Title] Research Report 30 Apr 2025 barney.sq.yao@htisec.com louis.yl.ng@htisec.com [Table_yemei1] Flash Analysis [Table_summary] 事件： 2025 年 4 月 28 日，阿里正式发布了新一代 Qwen-3 系列大语言模型（LLMs），包括从百亿参数到数十亿参数多个 量级的模型版本。2025 年 4 月 30 日，小米正式发布并开源了其首个专为推理任务设计的大语言模型——Xiaomi MiMo；DeepSeek 团队在 Hugging Face 平台发布了其最新的大语言模型——DeepSeek-Prover-V2-671B 点评： 阿里 Qwen-3： Qwen-3 系列具 ...

Domestic News - The manufacturing Purchasing Managers' Index (PMI) for April is reported at 49.0%, a decrease of 1.5 percentage points from the previous month, indicating a decline in manufacturing activity [3] - The new Private Economy Promotion Law will come into effect on May 20, aiming to support the development of the private sector [4] - The total holdings of gold ETFs in the Chinese market have reached a historical high, as reported by the World Gold Council [6] - The People's Bank of China conducted a 12 billion yuan reverse repurchase operation using a fixed quantity and interest rate bidding method [11] International News - Traders are fully pricing in four rate cuts of 25 basis points by the Federal Reserve by the end of 2025 [1] - Global gold demand in Q1 reached the highest level for a first quarter since 2016, according to the World Gold Council [2] - The U.S. economy has contracted, with a reported GDP decline of 0.3% in the first quarter, marking the first economic shrinkage since 2022 [6]

就在所有人都在期待DeepSeek官宣R2大模型之际，公司却出其不意地在"五一"前夕投下了另一枚技术炸弹。 4月30日，DeepSeek在Hugging Face平台上悄然开源了其最新模型——DeepSeek-Prover-V2-671B，一个专注于数学定理证明的大语言模型，专门针 对形式化数学证明任务进行优化。 DeepSeek-Prover-V2-671B使用了DeepSeek-V3架构，参数高达6710亿，采用MoE（混合专家）模式，具有61层Transformer层，7168维隐藏层。 | Hugging Face Q. Search models, datasets, users ... | | Models | ■ Datasets ■ Spaces Posts | Docs | Enterprise | Pricing | VII | Log In Sign Up | | --- | --- | --- | --- | --- | --- | --- | --- | --- | | < deepseek-ai/DeepSeek-Prover-V2-671B = 0 Wke 152 | Follo ...