上周品牌工程指数多只成分股表现强势。具体来说，兆易创新上涨11.23%，排在涨幅榜首位；中际旭 创上涨10.88%，居次席；北方华创、中兴通讯涨逾10%；阳光电源、安集科技涨逾8%；丸美生物和安 琪酵母分别上涨7.89%和5.40%；我武生物、迈瑞医疗、国瓷材料、恒瑞医药、澜起科技涨逾4%；广联 达、紫光股份、恒生电子、豪威集团、以岭药业、长电科技涨逾3%；老板电器、信立泰、泰格医药、 华谊集团、盾安环境等多只成分股涨逾2%。 □本报记者 王宇露 上周市场反弹，中证新华社民族品牌工程指数上涨2.08%，报1978.41点。兆易创新、中际旭创、北方华 创等成分股上周表现强势；下半年以来，中际旭创、阳光电源、兆易创新等成分股涨幅居前。 展望后市，机构认为，短期不确定性因素减少，下行风险释放较为充分，后续股市可能表现为震荡修 复。从中长期视角看，此轮大行情的基础并未发生改变。 多只成分股表现强势 上周市场反弹，上证指数上涨1.40%，深证成指上涨3.56%，创业板指上涨4.54%，沪深300指数上涨 1.64%，品牌工程指数上涨2.08%，报1978.41点。 从中长期视角看，星石投资认为，此轮大行情的基础并未发生 ...

Core Insights - The mobile industry is entering a new chapter with the upcoming release of an AI-native phone developed by ByteDance and ZTE, set to launch in early December [1][2] - This phone features a high-permission AI agent that integrates deeply with hardware, software, and the operating system, offering a significantly enhanced user experience compared to current AI smartphones [1][4] - The first generation of this AI phone has a production volume of approximately 30,000 units, with a second generation already in the planning stages for release in the first half of next year [1][2] Group 1: Industry Dynamics - The AI phone is positioned as a key player in the evolving landscape of mobile technology, with 2024 being regarded as the "year of AI phones" and expectations for killer AI applications by 2026 [4][5] - The competitive landscape for AI phones is emerging in three tiers: pioneers (Honor, OPPO, Huawei), ecosystem collaborators (Xiaomi, Vivo), and cross-industry entrants (ByteDance, ZTE Nubia) [4][5] - ByteDance's entry into the mobile market introduces new variables, leveraging its strengths in recommendation algorithms and natural language processing to potentially redefine user interaction from "clicking icons" to "intent recognition" [5][6] Group 2: Strategic Partnerships - ByteDance's collaboration with ZTE Nubia allows for a comprehensive approach to hardware design and manufacturing, addressing ByteDance's previous shortcomings in hardware implementation [2][3] - The AI phone project is part of ByteDance's broader strategy to explore new user interface interactions and enhance its AI hardware capabilities, with the Ocean team responsible for multiple AI device developments [2][3] - ZTE has the opportunity to reshape its hardware and software capabilities through this partnership, potentially positioning itself among the top-tier brands in the AI era [5][7] Group 3: Future Outlook - The AI-native phone represents a culmination of ByteDance's efforts to establish a full-stack AI ecosystem, integrating computing power, model capabilities, applications, and hardware [3][6] - The collaboration signifies a shift in strategy for ByteDance, moving from merely being an app developer to a player in the hardware space, aiming to control the entry points of AI technology [6][7] - Both ByteDance and ZTE face challenges in the early stages of AI phone development, but they also share significant opportunities for growth and innovation in the market [7]

Market Overview - On November 28, the Shanghai Composite Index rose by 0.34%, the Shenzhen Component Index increased by 0.85%, the ChiNext Index went up by 0.70%, and the CSI 300 Index gained 0.25% [1] - Among the tradable A-shares, 4,122 stocks increased, accounting for 75.87%, while 1,193 stocks declined [1] Capital Flow - The main capital saw a net inflow of 10.84 billion yuan throughout the day [1] - The ChiNext had a net inflow of 467 million yuan, and the Sci-Tech Innovation Board saw a net inflow of 320 million yuan, while the CSI 300 component stocks experienced a net outflow of 1.38 billion yuan [1] Industry Performance - Out of the 29 first-level industries classified by Shenwan, the top-performing sectors were Steel and Agriculture, Forestry, Animal Husbandry, and Fishery, both with a rise of 1.59% [1] - The sectors with the largest declines were Banking and Coal, with decreases of 0.83% and 0.14%, respectively [1] Industry Capital Flow - A total of 19 industries experienced net inflows, with the Electronics sector leading at a net inflow of 3.423 billion yuan and a daily increase of 1.30% [1] - The Non-ferrous Metals sector followed with a daily increase of 1.44% and a net inflow of 3.051 billion yuan [1] - Conversely, 12 industries saw net outflows, with the Pharmaceutical and Biological sector leading at a net outflow of 2.334 billion yuan despite a slight increase of 0.14% [1] - The Communication sector had a net outflow of 1.311 billion yuan with a daily increase of 0.12% [1] Individual Stock Performance - A total of 2,360 stocks had net inflows, with 786 stocks seeing inflows exceeding 10 million yuan, and 115 stocks with inflows over 100 million yuan [2] - The stock with the highest net inflow was Aerospace Development, which rose by 10.02% with a net inflow of 1.073 billion yuan [2] - Other notable inflows included Shannon Chip Creation and Duofu Duo, with net inflows of 902 million yuan and 778 million yuan, respectively [2] - Conversely, 67 stocks experienced net outflows exceeding 100 million yuan, with the highest outflows from Zhongji Xuchuang, ZTE Corporation, and Zhongsheng Pharmaceutical, with outflows of 797 million yuan, 695 million yuan, and 684 million yuan, respectively [2]

2025.11. 28 本文字数：745，阅读时长大约2分钟 作者 | 一财阿驴 11月28日，A股三大指数低开高走，截至收盘，沪指涨0.34%，深成指涨0.85%，创业板指涨0.7%。沪指本月累计跌 1.67%，深成指本月累跌2.95%，创业板指累跌4.23%。 | 代码 | 名称 | 两日图 | 现价 | 涨跌 | 涨跌幅 | | --- | --- | --- | --- | --- | --- | | 000001 | 上证指数 | Min | 3888.60 | 13.34 | 0.34% | | 399001 | 深证成指 | mm | 12984.08 | 108.89 | 0.85% | | 399006 | 创业板指 | ma | 3052.59 | 21.29 | 0.70% | 盘面上，福建板块午后涨势扩大，锂电池产业链全线反弹，商业航天概念午后回升；银行、医药股逆势飘绿。 具体来看，商业航天概念集体爆发，航天发展、四川金顶、乾照光电、航天环宇等10股涨停。 | 代码 | 名称 | 涨幅量 | 现价 | | --- | --- | --- | --- | | 300102 | 乾照光电 ...

Core Insights - The article presents a new research study that challenges the traditional belief that supervised fine-tuning (SFT) is a necessary precursor to reinforcement learning (RL) in the training of multimodal models, demonstrating that RL alone can effectively optimize multimodal capabilities [2][36]. Group 1: Research Findings - The study, conducted by Central South University and ZTE Corporation, introduces a quantifiable and operational "difficulty sampling" standard for multimodal models, validating the effectiveness of a training approach that relies solely on RL strategies (GRPO) [3][36]. - The research addresses two long-standing issues in multimodal post-training: the lack of quantifiable sample difficulty metrics and the inability of training paradigms to optimize perception and reasoning capabilities simultaneously [4][5][6]. Group 2: Methodology - Two complementary difficulty quantification strategies are proposed: Progressive Image Semantic Masking (PISM) and Cross-Modality Attention Balance (CMAB), which facilitate the hierarchical training framework [7][36]. - PISM involves progressively masking different parts of images to simulate varying degrees of visual information loss, allowing for the assessment of model performance based on its reliance on visual details [10][14]. - CMAB evaluates the complexity of cross-modal interactions by analyzing the attention scores of generated tokens across different Transformer layers, providing insights into the balance of attention between text and image inputs [19][34]. Group 3: Experimental Results - The experimental results indicate that the GRPO-only paradigm, which utilizes medium and difficult samples, significantly outperforms both full dataset training and random sample training, underscoring the importance of data quality over quantity [29][36]. - In visual reasoning tasks, the GRPO-only approach achieved optimal scores in multiple metrics, with notable improvements in MathVista (68.3) and OCRBench (77.8) compared to traditional methods [27][29]. - The study also highlights that SFT did not contribute to performance gains, suggesting that it may introduce "pseudo chains of thought" that limit the model's true reasoning capabilities [29][36]. Group 4: Future Directions - The research team outlines three future research directions: dynamic difficulty adjustment for adaptive learning, exploration of combined sampling strategies from PISM and CMAB, and validation of methods on larger multimodal models [38][39].

Core Points - Huawei and ZTE have signed 5G agreements in Vietnam, indicating a warming relationship between Vietnam and China amid deteriorating ties with the U.S. due to tariffs [1][2] - Vietnam has historically been cautious about using Chinese technology in sensitive infrastructure, but recent developments suggest a shift in this stance [1][4] - Contracts awarded to Huawei and ZTE include a $23 million deal for 5G equipment and additional contracts exceeding $20 million for 5G antennas [1][5] Group 1: Contracts and Agreements - Huawei and ZTE have secured multiple contracts for supplying 5G equipment in Vietnam, with Huawei winning a $23 million contract in April and ZTE winning contracts totaling over $20 million recently [1][5] - Despite concerns from Western officials, the contracts reflect Vietnam's changing approach towards Chinese technology amid geopolitical tensions [1][2] Group 2: Geopolitical Context - The U.S. has pressured Vietnam to avoid Chinese contractors in its digital infrastructure projects, including undersea cables, citing national security risks [4][6] - Recent meetings among Western officials in Vietnam have raised concerns about the implications of using Chinese technology on trust in Vietnam's networks and access to advanced U.S. technology [6][7] Group 3: Industry Dynamics - Huawei and ZTE face restrictions in the U.S. and some European countries due to perceived national security risks, yet they continue to expand their presence in Vietnam [5][9] - The European Commission is exploring measures to compel member states to exclude high-risk suppliers like Huawei and ZTE from telecom networks, which could lead to political disputes [7][9]

Group 1 - A total of 3735 stocks in the A-share market received financing purchases on November 27, with 325 stocks having purchase amounts exceeding 100 million yuan [1] - The top three stocks by financing purchase amount were Zhongji Xuchuang, Xinyi Sheng, and Shenghong Technology, with amounts of 4.428 billion yuan, 3.497 billion yuan, and 2.077 billion yuan respectively [1] - Four stocks had financing purchase amounts accounting for over 30% of the total transaction amount, with Jicheng Electronics, Tongling Shares, and Wuchan Zhongda leading at 35.54%, 31.53%, and 30.87% respectively [1] Group 2 - There were 24 stocks with net financing purchases exceeding 100 million yuan, with Shenghong Technology, Hanwujishi-U, and ZTE ranking the highest at 501 million yuan, 370 million yuan, and 286 million yuan respectively [1]

Core Viewpoint - China's leading telecommunication firms, Huawei and ZTE, have secured multiple contracts this year to supply 5G equipment in Vietnam, indicating a strengthening relationship between Hanoi and Beijing, which raises concerns among Western officials [1] Group 1 - Huawei and ZTE are the primary companies involved in the 5G equipment supply contracts in Vietnam [1] - The contracts signify a deepening of ties between Vietnam and China, reflecting geopolitical shifts in the region [1] - The involvement of these Chinese firms has raised alarms among Western officials regarding security and influence in Southeast Asia [1]

Group 1 - The total market value of listed companies in Guangdong Province has surpassed 30 trillion yuan, with a net increase of over 14 trillion yuan during the "14th Five-Year Plan" period, equivalent to the province's GDP in 2024 [1][4] - As of November 18, 2023, there are 887 A-share listed companies in Guangdong, with a total market value reaching 19 trillion yuan [1][3] - Guangdong has added 255 new listed companies, with significant contributions from the Sci-Tech Innovation Board, Growth Enterprise Market, and Beijing Stock Exchange [1][5] Group 2 - Guangdong's listed companies have effectively responded to major risks and challenges, achieving steady growth in quantity and quality, showcasing their role as economic stabilizers and innovation leaders [6][10] - The number of listed companies with a market value exceeding 1 trillion yuan has increased from one (Tencent) to four, with 25 companies valued over 100 billion yuan [6][10] Group 3 - The number of high-tech enterprises among Guangdong's listed companies reached 628, accounting for 71%, the highest in the country [14] - Research and development personnel in Guangdong's listed companies exceeded 750,000, with a compound annual growth rate of R&D investment reaching 12% [14][15] Group 4 - Guangdong's listed companies have completed 123 mergers and acquisitions, leading the nation in both the number of projects and transaction amounts [17][21] - The province has seen a significant increase in overseas listings, with 322 companies listed abroad, raising 480.28 billion yuan through IPOs [21][22] Group 5 - Guangdong's listed companies are increasingly focusing on internationalization, with a notable rise in overseas revenue from 0.96 trillion yuan in 2020 to 1.86 trillion yuan in 2024 [21][22] - The province is positioned to leverage opportunities in AI technology and aims to build a resilient and self-controlled industrial system [23]

Group 1 - Citigroup's stake in ZTE Corporation's H-shares decreased from 8.92% to 8.62% as of November 21 [1]