Core Points - DeepSeek has issued a statement regarding fraudulent activities where individuals impersonate the company or its employees, using forged identification and business licenses to scam users under the guise of "computing power leasing" and "equity financing" [1][2][3] - The fraudulent actions have severely harmed user rights and damaged the company's reputation [1][2] Company Policy - DeepSeek has never requested users to make payments to personal or unofficial accounts; any such requests for private transfers are considered scams [3] - Any activities that misuse the company's name for "computing power leasing" or financing are illegal, and the company will pursue legal action against such actions [3] User Advisory - Users are advised to obtain official information and updates through the official website (deepseek.com) and verified accounts [1] - The company's official webpage and app products are currently free; for API services, users should recharge through the official platform, with the official payment account name being "Hangzhou DeepSeek Artificial Intelligence Basic Technology Research Co., Ltd." [1] - In case of suspicious situations, users should verify through the official email or report to law enforcement [1]

Core Insights - The DeepSeek-R1 inference model research paper has made history by being the first Chinese large model research to be published on the cover of the prestigious journal Nature, marking a significant recognition of China's AI technology in the international scientific community [1][2] - Nature's editorial highlighted that DeepSeek has broken the gap of independent peer review for mainstream large models, which has been lacking in the industry [2] Group 1: Research and Development - The DeepSeek-R1 model's research paper underwent a rigorous peer review process involving eight external experts over six months, emphasizing the importance of transparency and reproducibility in AI model development [2] - The paper disclosed significant details about the training costs and methodologies, including a total training cost of $294,000 (approximately 2.09 million RMB) for R1, achieved using 512 H800 GPUs over 198 hours [3] Group 2: Model Performance and Criticism - DeepSeek addressed initial criticisms regarding the "distillation" method used in R1, clarifying that all training data was sourced from the internet without intentional use of outputs from proprietary models like OpenAI's [3] - The R1 model has been recognized for its cost-effectiveness compared to other inference models, which often incur training costs in the tens of millions [3] Group 3: Future Developments - There is significant anticipation regarding the release of the R2 model, with speculation that delays may be due to computational limitations [4] - The recent release of DeepSeek-V3.1 has introduced a mixed inference architecture and improved efficiency, indicating a step towards the "Agent" era in AI [4][5] - DeepSeek's emphasis on using UE8M0 FP8 Scale parameter precision in V3.1 suggests a strategic alignment with domestic AI chip development, potentially enhancing the performance of future models [5]

Core Insights - DeepSeek's research paper, "DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning," has been published in the prestigious journal Nature, highlighting significant advancements in AI reasoning capabilities [1][11]. Group 1: Research and Development - The initial version of DeepSeek's paper was released on arXiv in January, and the Nature publication includes more detailed model specifications and reduced anthropomorphism in descriptions [5]. - DeepSeek-R1's training cost was reported to be $294,000, with specific costs for different components outlined, including $202,000 for DeepSeek-R1-Zero training and $82,000 for SFT data creation [9]. - The training utilized A100 GPUs for smaller models and expanded to 660 billion parameters for the R1 model, demonstrating a scalable approach to model development [8][10]. Group 2: Model Performance and Validation - DeepSeek-R1 has become the most popular open-source inference model globally, with over 10.9 million downloads on Hugging Face, marking it as the first mainstream large language model to undergo peer review [11]. - The research emphasizes that significant reasoning capabilities can be achieved through reinforcement learning without relying heavily on supervised fine-tuning, which is a departure from traditional methods [13]. - The model's training involved a reward mechanism that encourages correct reasoning, allowing it to self-validate and improve its performance on complex tasks [13]. Group 3: Industry Implications - The findings from DeepSeek's research could set a precedent for future AI model development, particularly in enhancing reasoning capabilities without extensive data requirements [11][13]. - The independent peer review process adds credibility to the model's performance claims, addressing concerns about potential manipulation in AI benchmarking [11].

Market Performance - As of September 17, the Shanghai Composite Index rose by 0.37% to close at 3876.34 points, the Shenzhen Component Index increased by 1.16% to 13215.46 points, and the ChiNext Index gained 1.95% to 3147.35 points [1] - The Kweichow Moutai Semiconductor ETF (588170) increased by 3.64%, while the Semiconductor Materials ETF (562590) rose by 3.32% [1] - In the overnight U.S. market, the Dow Jones Industrial Average increased by 0.57%, while the Nasdaq Composite Index fell by 0.33% and the S&P 500 Index decreased by 0.10% [1] Industry Insights - The DeepSeek-R1 inference model research paper, led by Liang Wenfeng, was published in the prestigious journal Nature, marking it as the first mainstream large language model to undergo peer review [2] - Huawei released the "Smart World 2035" series of reports, predicting a significant increase in total computing power by 2035, with a 500-fold increase in AI storage capacity demand compared to 2025 [2] - The Zhangjiang Artificial Intelligence Innovation Town in Shanghai aims to gather over 500 AI companies by 2027 and achieve a scale of 100 billion yuan by 2030, supported by a 2 billion yuan fund initiated by Hillhouse Capital and local state-owned enterprises [3] - Tianfeng Securities anticipates a structural prosperity in the global semiconductor industry driven by rapid growth in AI computing demand, accelerated terminal intelligence, and the recovery of automotive electronics [3] Related ETFs - The Kweichow Moutai Semiconductor ETF (588170) tracks the Shanghai Stock Exchange's semiconductor materials and equipment index, focusing on semiconductor equipment (59%) and materials (25%) [4] - The Semiconductor Materials ETF (562590) also emphasizes semiconductor equipment (59%) and materials (24%), benefiting from the expansion of semiconductor demand driven by the AI revolution [4]

Core Insights - DeepSeek, a Chinese company focused on large language models (LLM) and artificial general intelligence (AGI), has gained attention for its open-source AI model DeepSeek-R1, which employs a large-scale inference model training method [1] - The training method was published in the prestigious journal Nature, revealing that the reasoning capabilities of LLMs can be enhanced through pure reinforcement learning, thereby reducing the human input required for performance enhancement [1] - The model outperformed traditional LLMs in tasks related to mathematics, programming competitions, and graduate-level STEM problems [1] Group 1 - DeepSeek-R1 includes a supervised in-depth training phase to optimize the reasoning process, utilizing reinforcement learning instead of human examples to develop reasoning steps, which reduces training costs and complexity [2] - The model achieved scores of 77.9% and 79.8% in mathematical benchmark tests for DeepSeek-R1-Zero and DeepSeek-R1, respectively, and also excelled in programming competitions and graduate-level biology, physics, and chemistry problems [2] - A concurrent article in Nature highlighted some limitations of the current version of DeepSeek-R1, such as language mixing and sensitivity to prompt engineering, indicating areas for improvement in future versions [2] Group 2 - The DeepSeek-AI team concluded that future research should focus on optimizing the reward process to ensure reliable reasoning and task outcomes [3]

Core Insights - DeepSeek's R1 model has been recognized as the first major language model to be peer-reviewed and published in the prestigious journal Nature, marking a significant milestone in AI research [1][2] - The R1 model achieved over 10.9 million downloads on Hugging Face, making it the most popular open-source inference model globally [2] - DeepSeek's innovative approach utilizes pure reinforcement learning to enhance reasoning capabilities, diverging from traditional human-imitation methods [2][3] Company Developments - DeepSeek's R1 model was developed with a training cost of only $294,000, significantly lower than the costs associated with training AI models by OpenAI and Google, which can reach millions [2] - The company released an upgraded version, DeepSeek-V3.1, which features a mixed reasoning architecture, improved thinking efficiency, and enhanced agent capabilities [3] - DeepSeek was founded in 2023 in Hangzhou, backed by the quantitative firm Huansquare, with a team composed of experts from top universities and international institutions [3] Industry Context - The publication of DeepSeek's research is seen as a critical step in addressing the rampant speculation and unverified claims within the AI industry, emphasizing the importance of independent peer review [3] - The recognition of DeepSeek's work by Nature highlights China's advancements in foundational research in large models, contributing to the global AI landscape [2]

Group 1 - DeepSeek-R1 reasoning model research paper, led by Liang Wenfeng, has been published in the prestigious journal Nature, marking it as the first mainstream large language model to undergo peer review [1] - The latest paper provides more details on model training and addresses initial concerns regarding model distillation, highlighting the significance of independent peer review in the AI field [1] - The AI industry is experiencing a positive cycle driven by performance and capital expenditure, with the domestic AI ecosystem rapidly developing across various segments including large models, computing power, and applications [1] Group 2 - As of September 18, 2025, the CSI Artificial Intelligence Theme Index (930713) rose by 0.65%, with notable gains from stocks such as Jingsheng Electronics (up 9.99%) and Rockchip (up 5.82%) [2] - The AI Artificial Intelligence ETF (512930) also increased by 0.66%, achieving a three-day consecutive rise, with a latest price of 2.13 yuan and a weekly increase of 8.08% [2] - The management fee for the AI Artificial Intelligence ETF is 0.15%, and the custody fee is 0.05%, making it the lowest among comparable funds, while it has the highest tracking accuracy of 0.008% over the past three months [2] Group 3 - As of August 29, 2025, the top ten weighted stocks in the CSI Artificial Intelligence Theme Index accounted for 60.82% of the index, with companies like Xinyi Technology and Zhongji Xuchuang leading the list [3] - The top ten stocks include Xinyi Technology (300502), Zhongji Xuchuang (300308), and Cambricon (688256), among others, indicating a concentration of investment in these key players within the AI sector [3][5]

开源人工智能（AI）的价值正获得更广泛的认可。 刚刚，DeepSeek-R1 论文以封面文章的形式登上了权威科学期刊 Nature，DeepSeek 创始人兼 CEO 梁文峰为该论文的通讯作者。 论文链接：https://www.nature.com/articles/s41586-025-09422-z 研究团队假设，人类定义的推理模式可能会限制模型的探索，而无限制的强化学习（RL）训练可以更好地激励大语言模型（LLM）中新推理能力的涌 现。 他们通过实验证明，LLM 的推理能力可以通过纯 RL 来提升，从而减少增强性能所需的人类输入工作量，且在数学、编程竞赛和 STEM 领域研究生水平 问题等任务上，比经传统方法训练的 LLM 表现更好。 DeepSeek-R1 推出后，得到了全球开发者的广泛好评，截至发文前，其在 GitHub 上的 star 数已经达到了 91.1k。 在一篇同期发表的观点与评论文章中，卡内基梅隆大学助理教授Daphne Ippolito和他的博士生张益铭（现为 Anthropic 的 LLM 安全和对齐研究员）评价 道： "DeepSeek-R1 已从一个强大但不透明的解决方案寻找者 ...

刚刚，DeepSeek-R1登上了Nature封面！ 今年1月，DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning论文发布，如今成功登上全球顶刊封面。 通讯作者梁文锋带队，用RL为大模型推理能力开辟了全新路径。 论文地址：https://www.nature.com/articles/s41586-025-09422-z 值得一的是，补充材料首次公开了R1训练成本——294000美元，数字低到惊人。 即便是加上约600万美元的基础模型成本，也远低于OpenAI、谷歌训练AI的成本。 在封面推荐中，Nature毫不吝啬地赞扬了DeepSeek-R1的成就。 开源之后，R1在Hugging Face成为最受欢迎的模型，下载量破1090万次。关键是，它是全球首个经过同行评审的主流大模型。 | Training Costs | DeepSeek-R1-Zero | SFT data creation | DeepSeek-R1 | Total | | --- | --- | --- | --- | --- ...

太令人意外！ 却又实至名归！ 最新一期的 Nature 封面，竟然是 DeepSeek-R1 的研究。 也就是今年 1 月份 DeepSeek 在 arxiv 公布的论文《DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning》。这篇Nature论文 通讯作者正是梁文锋。 论文链接： https://www.nature.com/articles/s41586-025-09422-z 在封面的推荐介绍中，Nature 写到： 如果训练出的大模型能够规划解决问题所需的步骤，那么它们往往能够更好地解决问题。这种『推理』与人类处理更复杂问题的方式类似，但 这对人工智能有极大挑战，需要人工干预来添加标签和注释。在本周的期刊中，DeepSeek 的研究人员揭示了他们如何能够在极少的人工输入 下训练一个模型，并使其进行推理。 DeepSeek-R1 模型采用强化学习进行训练。在这种学习中，模型正确解答数学问题时会获得高分奖励，答错则会受到惩罚。结果，它学会了推 理——逐步解决问题并揭示这些步骤——更有可能得出正确 ...