Core Points - The article discusses President Carter's unprecedented use of his pardon power, highlighting his record compared to other presidents [1] Group 1 - President Carter is noted for his extensive use of the pardon power, which is described as unparalleled among U.S. presidents [1]

打打谈谈近10年，贸易战的决战序幕终于在10月份拉开，强硬的特朗普突然表达愿意和我们进行理性谈 判，还暗示百分百关税纯口嗨，最后不一定会落实，反转确实来得猝不及防，那么特朗普为什么会怂得 这么快？ ...

Core Viewpoint - The film "Loveable" explores the emotional struggles of the protagonist Maria, who faces challenges in her second marriage and parenting, ultimately leading to her awakening from a victim mentality to taking responsibility for her life [1][5][51]. Group 1: Character Development - Maria's character embodies the struggles of many individuals, showcasing her journey from feeling like a victim to recognizing her own role in her circumstances [4][5]. - The film highlights the internal growth of Maria as she confronts her pain and learns to love herself, moving away from a narrative of victimhood [39][41][53]. Group 2: Psychological Themes - The concept of "victim narrative" is central to the film, illustrating how individuals can become trapped in a cycle of blame and helplessness, often stemming from their upbringing and societal structures [8][13][50]. - The film critiques the tendency to view oneself solely as a victim, emphasizing the importance of recognizing personal responsibility in relationships [30][46]. Group 3: Relationship Dynamics - Maria's interactions with her husband and children reveal a pattern of emotional manipulation, where she uses guilt to maintain control and validate her feelings of inadequacy [37][30]. - The film portrays the complexities of familial relationships, particularly how generational patterns of communication can perpetuate feelings of victimization and emotional distance [20][21][22]. Group 4: Conclusion and Reflection - The ending of the film suggests a shift towards self-acceptance and the acknowledgment of one's vulnerabilities, as Maria learns to express her fears and desires openly [53][54]. - The narrative encourages viewers to confront their own victim narratives and consider the possibility of transformation through self-awareness and responsibility [51][52].

Core Insights - The article discusses the challenges faced by current LLMs in processing long texts due to quadratic growth in computational complexity, which increases with longer sequences [1] - DeepSeek-OCR presents a novel approach to address this issue by utilizing "optical compression," converting text into images to reduce the number of tokens required for processing [5][34] - The model demonstrates a token compression capability of 7 to 20 times while maintaining high accuracy, showcasing its potential for efficient long-context processing [34][36] Technical Overview - DeepSeek-OCR achieves a compression rate of up to 10 times with an accuracy of over 97% [4] - The model uses a two-component architecture: DeepEncoder for image feature extraction and compression, and DeepSeek3B-MoE for reconstructing text from compressed visual tokens [16][18] - The DeepEncoder employs a clever architecture combining SAM-base and CLIP-large models, along with a convolutional compressor to significantly reduce token numbers before entering the global attention layer [10][11] Performance Metrics - OmniDocBench benchmark results indicate that a single A100-40G GPU can generate over 200,000 pages of LLM/VLM training data daily, while 20 nodes (160 A100 GPUs) can produce up to 33 million pages [7] - DeepSeek-OCR outperforms existing models, requiring only 100 visual tokens to exceed the performance of GOT-OCR2.0, which uses 256 tokens per page [15] Data Utilization - The DeepSeek team collected 30 million pages of multilingual PDF data, covering around 100 languages, with a focus on Chinese and English [21] - The data is categorized into coarse and fine annotations, with high-quality data generated through various models to enhance recognition capabilities [22] Application Potential - DeepSeek-OCR not only recognizes text but also possesses deep parsing capabilities, making it suitable for STEM applications that require structured extraction from complex images [27] - The model can extract structured data from financial reports, chemical structures, geometric figures, and generate dense captions for natural images [28] Future Directions - The team proposes exploring the concept of "optical compression" to simulate human memory decay, allowing for efficient processing of long contexts by reducing the fidelity of older information [30][31] - Future plans include conducting systematic evaluations and pre-training methods to further validate the effectiveness of this approach [35]

美股大跌，6次谈崩，政府停摆！懂王又在搞新花活了？关税帝君烽火戏诸侯，一口气看懂，美国政府 怎么突然关门了？ ...

Core Insights - Jay Chou's recent troubles involve a Bitcoin account managed by his magician friend, Cai Weize, who claimed the account was locked a year ago, resulting in a loss of funds [1][2] - The article discusses the emergence of AI models competing in the cryptocurrency market, highlighting a competition called Alpha Arena where six top AI models are trading cryptocurrencies [3][4] Group 1: AI Competition Overview - The competition involves six AI models, each given $10,000 to trade perpetual contracts on the Hyperliquid platform, with trading pairs including BTC, ETH, BNB, SOL, XRP, and DOGE [4][6] - The performance of these AI models is measured by risk-adjusted returns, focusing not only on profits but also on the risks taken [6][7] Group 2: Performance of AI Models - As of the latest update, DeepSeek Chat V3.1 leads with an account value of $14,310 and a return of +43.1%, showcasing a strategy of high leverage and concentrated positions [11][12] - Grok 4 follows with an account value of $13,921 and a return of +39.21%, employing a high-leverage long-only strategy [12][21] - Claude Sonnet 4.5 has an account value of $12,528 and a return of +25.28%, focusing on a conservative trading approach [12][23] - In contrast, GPT-5 and Gemini 2.5 Pro are underperforming, with returns of -24.78% and -27.74% respectively, indicating poor trading strategies and high transaction costs [12][30] Group 3: AI's Role in Investment - The article emphasizes that AI's greatest value in investment may lie in transparency, allowing investors to see trading records and decision-making processes, unlike human-managed accounts [40][41] - The ambition behind the AI competition is to use financial markets as a training ground for AI, aiming for continuous learning and adaptation to market dynamics [34][35]

Core Insights - The rapid growth of Sam's Club's instant retail business has led to concerns about delivery riders facing overloading issues rather than speeding [1][4] - Sam's Club has opened four new "cloud warehouses" across various cities, enhancing its "one-hour express delivery" service [2][3] - The increase in cloud warehouses has raised questions about the safety and rationality of Sam's delivery methods due to reports of overloaded delivery vehicles [4][6] Group 1: Business Expansion - Sam's Club has been expanding its cloud warehouse network, with at least seven new warehouses added in the past month, bringing the total close to 500 [3][30] - The "express delivery" service is not self-operated but contracted to third-party platforms, which has implications for rider compensation and workload [11][12] - The cloud warehouse model allows Sam's Club to test market demand before opening new physical stores, with a strategy of "warehouse first, store later" [26][31] Group 2: Delivery Challenges - Delivery riders are incentivized to take on more orders to increase their income, leading to a culture of overloading vehicles with heavy goods [5][18] - The average weight limit for deliveries is set at 30 kilograms, but riders often exceed this limit due to operational loopholes [16][20] - Riders report that their income has not significantly increased despite higher order volumes, leading to dissatisfaction and potential safety risks [19][20][35] Group 3: Financial Performance - Walmart China reported a net sales figure of $12.365 billion for the first half of the 2026 fiscal year, with a year-on-year increase of $2.482 billion [34] - The second quarter saw net sales of $5.786 billion, reflecting a growth rate of 30.8%, with online sales growing at a rate of 39% [34][32] - The success of the cloud warehouse and express delivery model in China is being studied for application in other global markets [33]

Core Insights - Leapmotor's flagship D platform aims to redefine industry standards with a million-level experience and several pioneering technologies [1] Group 1 - The focus is on the core technologies of Leapmotor, highlighting their advanced features [1]

Group 1 - The core viewpoint of the articles highlights the significant growth and global dominance of China's manufacturing sector, which has seen its value-added manufacturing increase from 26.6 trillion yuan to 33.6 trillion yuan from 2020 to 2024, contributing over 30% to global manufacturing growth during the 14th Five-Year Plan period [2][3][4] - China's manufacturing value-added accounted for approximately 30% of the global total, maintaining the largest share for 15 consecutive years, with projections indicating it could rise to 45% by 2030 [6][9] - The manufacturing sector's output is primarily driven by domestic demand, with less than 30% of production being exported, indicating a strong internal market [14][15] Group 2 - The automotive and semiconductor industries are identified as key areas for growth, with China's automotive production expected to reach 31.28 million units in 2024, accounting for 33.8% of global output [10][11] - Despite the strong performance in manufacturing, challenges remain in specific sectors such as semiconductors, where China faces significant trade deficits, highlighting the need for improvement in these critical areas [12][18] - The articles emphasize the importance of China's manufacturing capabilities in supporting various sectors, including agriculture and services, and the potential for further development in the third industry [58][59] Group 3 - The articles discuss the implications of China's manufacturing strength on global trade dynamics, noting that China's trade surplus has reached unprecedented levels, significantly impacting the global economy [17][30] - The manufacturing sector's ability to adapt and respond to global demands is underscored, with the potential for continued expansion in international markets, particularly in developing regions [22][25] - The articles also highlight the increasing internationalization of the renminbi, driven by China's manufacturing exports, which is reshaping global payment systems [31][33] Group 4 - The articles point out the internal challenges within China's manufacturing sector, including issues related to overcapacity and the need for regulatory oversight to ensure fair competition [54][56] - The manufacturing industry's employment impact is significant, with approximately 1.3 billion people employed in this sector, underscoring its role in the broader economy [56] - The articles conclude that while China's manufacturing sector has achieved remarkable growth, it must navigate both domestic and international challenges to sustain its competitive edge [58][59]

Group 1 - The core point of the article is that the recent rebound in the technology sector is primarily influenced by the attitude of capital, with macro data remaining acceptable and policy resources shifting towards industrial upgrades and technological breakthroughs [1][3] - The domestic market has stabilized, particularly with a short-term rebound in the technology sector, driven by positive performance in communication equipment, especially optical modules, and a recovery in artificial intelligence and robotics [3] - There are rumors that leading companies may secure large overseas orders, significantly revising revenue and profit expectations for 2026, although these claims have not been officially confirmed [3] Group 2 - The easing of external environmental risks, particularly the softening of US-China relations, is seen as a key driver for the market recovery rather than the industry developments themselves [3] - Recent statements from the US indicate that the focus of upcoming US-China negotiations will include three main topics: China's rare earth regulations, the purchase of US soybeans, and the control of addictive substances like fentanyl, suggesting potential for substantive outcomes in the near future [4] - The upcoming summit in South Korea is anticipated to yield positive news, contributing to improved market sentiment [4]