Core Viewpoint - The article discusses the innovative framework DrSR (Dual Reasoning Symbolic Regression) developed by researchers at the Institute of Automation, Chinese Academy of Sciences, which enables large models to analyze data, reflect on failures, and optimize models like scientists do [2][14][56]. Group 1: Framework and Mechanism - DrSR employs a dual-path reasoning mechanism that integrates "data insights" and "experience summaries" to guide large models in scientific equation discovery [16][28]. - The framework consists of three virtual scientists: a data scientist, a theoretical scientist, and an experimental scientist, each contributing to a collaborative mechanism for efficient scientific equation discovery [3][7]. Group 2: Performance and Results - In various interdisciplinary modeling tasks, DrSR has demonstrated superior generalization capabilities, outperforming existing methods in accuracy and efficiency [4][30]. - Experimental results show that DrSR achieved an accuracy of 99.94% in nonlinear damping oscillation system modeling, significantly surpassing all baseline methods [31]. Group 3: Learning and Adaptation - DrSR's process is a closed loop: data analysis → prompt guidance → equation generation → evaluation and scoring → experience summarization, allowing the model to accumulate knowledge and refine its approach [28]. - The framework's experience-driven strategy helps avoid common failure structures, resulting in a higher proportion of valid equations generated compared to other methods [37]. Group 4: Robustness and Generalization - DrSR exhibits strong robustness against noise and out-of-distribution (OOD) data, maintaining low normalized mean square error (NMSE) across various tasks [40][41]. - The model's performance remains stable under different Gaussian noise levels, showcasing its generalization advantages [41]. Group 5: Future Directions - DrSR is integrated into the ScienceOne platform, providing efficient and interpretable scientific modeling services, with plans to enhance its reasoning capabilities and cross-task generalization [57]. - Future improvements will focus on expanding DrSR's capabilities to multi-modal scientific modeling scenarios and incorporating continuous learning mechanisms [61].

Group 1 - The event "IQ Talk" focused on the development strategies, key technologies, industrial applications, and talent cultivation in the field of scientific intelligence, highlighting the transformative impact of AI on scientific research [1][2] - Experts believe that "AI for Science" signifies a paradigm shift in scientific research, enabling a new form of interdisciplinary collaboration and innovation [1] - The need for new talent with cross-disciplinary thinking, engineering practice, and social insight was emphasized to navigate the scientific cognitive revolution [1] Group 2 - AI-driven technologies are expected to overcome traditional R&D challenges such as long cycles and high costs, facilitating new scientific discoveries [2] - AI tools are being developed to address real scientific problems, including algorithms for predicting biological interactions and driving drug molecule design [2] - Demonstrations of AI platforms that integrate literature review, computation, experimentation, and multidisciplinary collaboration were showcased, indicating a shift from trial-and-error to computation-driven research paradigms [2]

Core Viewpoint - The "Science Intelligence White Paper 2025" emphasizes the integration of AI in scientific instruments, transforming them from mere data collection tools to intelligent research partners, enhancing efficiency and capabilities across various scientific fields [1][4][5]. Group 1: Integration of AI in Scientific Instruments - The white paper highlights the deep integration of scientific intelligence (AI for Science, AI4S) with scientific instruments, showcasing how AI empowers traditional instruments to become intelligent research partners [5]. - AI-driven automation in experiments has shown significant improvements, such as a 30% increase in stability and a 50% boost in data collection efficiency in nuclear fusion research [5][6]. - In drug development, automated laboratories utilizing AI algorithms can complete thousands of compound screenings within 48 hours, achieving speeds 10 times faster than traditional methods [6]. Group 2: Trends in Scientific Instrument Intelligence - The design philosophy of scientific instruments is undergoing three major transformations: - Human-machine interaction revolution, with 90% of new instruments featuring smart interfaces, including gesture control [11]. - Integrated data management and analysis, allowing instruments to generate visual reports directly, such as automatic mutation annotation in gene sequencers [13]. - Sustainable design, with 60% of instruments expected to use bio-based resins or recycled metals [15]. Group 3: Global Competitive Landscape and China's Breakthroughs - China is leading in specific fields such as Earth and environmental sciences, with advancements in AI meteorological models and remote sensing instruments [17]. - The domestic production rate of online detection instruments in intelligent manufacturing has surpassed 60% [18]. - However, there are challenges, including a reliance on imports for high-end analytical instruments, with over 80% of such instruments being imported [19]. Group 4: Future Outlook for Scientific Instruments - The future of scientific instruments is projected to focus on three main directions: - Intelligent upgrades, with AI deeply embedded in instrument control and data analysis processes [23]. - Development of specialized instruments for extreme environments, such as deep-sea and space applications [24]. - Establishing an open ecosystem through global laboratory alliances to share material databases [25].

Group 1 - Shanghai's first multimodal large model in the transportation sector has been launched, expected to improve intersection traffic efficiency by approximately 15% [1] - Red Hat has announced the launch of the open-source project llm-d, aimed at meeting the large-scale inference needs of generative AI, in collaboration with contributors like NVIDIA and Google Cloud [1] - Fudan University has released "AI Big Class 2.0" and the 2.0 version of its intelligent computing platform, emphasizing scientific intelligence [1] Group 2 - OpenAI plans to release its first AI hardware in 2026, driven by ChatGPT, aiming to integrate AI more deeply into users' daily lives [1]

Group 1 - Fudan University officially launched the CFFF Intelligent Computing Platform 2.0, which has undergone comprehensive upgrades in scientific model openness, scientific data security sharing, and software-hardware collaborative optimization compared to CFFF 1.0 [1][3] - The CFFF platform is the largest cloud-based research intelligence platform in domestic universities, set to officially go live on June 27, 2023, and covers the entire "teaching-learning-research" chain in artificial intelligence infrastructure [1][3] - The platform features 47 specialized academic models across various disciplines and provides over 40,000 datasets totaling 11PB for scientific model development and secure sharing [1][3] Group 2 - Fudan University aims to continuously promote AI course development and establish CFFF 2.0 as a foundational research facility in the field of scientific intelligence, facilitating a reform in teaching and learning [3] - The university's president emphasized the platform's role in accelerating the discovery of new scientific principles and technological breakthroughs through high-quality scientific resources and research tools [3][4] - A global alliance for scientific intelligence universities was proposed to foster collaboration, resource sharing, and ecological development in the field of scientific intelligence [4][6] Group 3 - Fudan University released the "AI Big Class" 2.0 white paper, focusing on collaborative innovation between teachers and students, and established an AI education and teaching innovation center to support the Shanghai International Science and Technology Innovation Center [6][7] - A strategic partnership with Springer Nature was formed to launch a comprehensive academic journal titled "Science and AI," aimed at addressing complex challenges through artificial intelligence and scientific principles [7]

Group 1 - Fudan University and Shanghai Institute of Science and Intelligent Research jointly released the "Science Intelligence White Paper 2025" in collaboration with Springer Nature's Nature Research Intelligence [1]

Core Insights - The Chinese Academy of Sciences has launched an intelligent research platform called ScienceOne, which aims to enhance scientific research across various disciplines by leveraging a foundational scientific model [1][2]. Group 1: ScienceOne Overview - ScienceOne focuses on common scientific research needs across disciplines, achieving breakthroughs in data understanding, computational optimization, and reasoning evaluation [1]. - The platform acts as an AI research assistant, empowering various research processes such as hypothesis generation, experimental validation, and pattern discovery [1]. Group 2: Product Features - ScienceOne includes two main products: S1-Literature Literature Assistant and S1-ToolChain Scientific Tool Scheduler [2]. - S1-Literature can automatically generate high-level literature reviews and deeply understand scientific data types, allowing users to summarize thousands of papers with simple commands [2]. - S1-ToolChain enables autonomous collaboration of scientific tools for cross-disciplinary data understanding and scientific computation, integrating nearly 300 tools for various scientific analyses [3]. Group 3: Future Developments - The development team plans to open-source the foundational scientific model S1-Base and release a scientific AI factory S1-Agent, aiming to create a platform-based tool system [3].

Group 1 - Beijing's higher education admission process for undergraduate programs will allow students to fill out their applications for the first time from June 27 to July 1, 2025, following the release of their exam scores [1] - The Tongzhou District of Beijing has announced its 2025 compulsory education enrollment policy, which will primarily use a registration-based admission system, considering factors such as housing and residency [2] - The "Beijing University Student Entrepreneurship Park" has been officially launched, focusing on key industries such as high-end equipment and life health, to encourage student entrepreneurship [3] Group 2 - Peking University has established a new School of Science and Intelligence at its Shenzhen Graduate School, aimed at integrating artificial intelligence with basic sciences to cultivate interdisciplinary talents [4] Group 3 - Kede Education reported a 2.27% year-on-year decline in revenue for Q1 2025, totaling 183 million yuan, and a net profit of 37.97 million yuan, reflecting a 6.8% decrease compared to the previous year [5]

Core Insights - The article discusses the breakthrough of Shanghai's "Fuyiao" meteorological model, which aims to provide highly localized weather forecasts for urban areas, enhancing precision and response time in weather prediction [1][2][3]. Group 1: Technological Advancements - The "Fuyiao" model has improved spatial resolution by three times to 1 kilometer and reduced inference time from 10 minutes to 3 seconds, likening it to a "digital nervous system" for cities [3]. - The model integrates multi-modal data from radar, satellites, and numerical forecasts, achieving an end-to-end forecasting loop for short-term severe weather events [4]. - The CFFF (Computing for the Future at Fudan) platform, with a computing power of 40 PFlop/s, supports various scientific fields, including life sciences and atmospheric sciences, enhancing research capabilities [7][12]. Group 2: Collaborative Research and Development - The Shanghai Intelligent Research Institute (上智院) collaborates with universities and industries to drive innovation across multiple scientific domains, focusing on AI-driven research paradigms [4][12]. - The CFFF platform has linked nearly 4,000 users across 51 departments, supporting numerous high-impact research papers and national projects [12]. - The institute aims to break down barriers in vertical scientific fields by developing specialized AI models, such as the "Fuxi" meteorological model and the "Nüwa" life sciences model [14][16]. Group 3: Practical Applications - The "Fuxi" meteorological model has demonstrated its practical value by accurately predicting the landing point of Typhoon "Bebinca," outperforming other forecasting models [14]. - The "Nüwa" model focuses on drug development, particularly in siRNA research, significantly improving the accuracy of drug efficacy predictions [16][18]. - The "Suiren" molecular model addresses core challenges in material science, aiding in the development of sustainable materials and enhancing battery research [18][19]. Group 4: Climate and Social Sciences - The Planet Intelligence@Climate (PI@Climate) initiative aims to create a climate science language model that integrates data from various disciplines to address complex climate issues [22][23]. - The initiative has developed a benchmark system to validate the capabilities of climate models, enhancing their applicability in real-world scenarios [23][25]. - The collaboration between AI technology and humanities research is exemplified by the "Chinese Civilization Model," which seeks to analyze human interactions and cultural heritage [29][30]. Group 5: Infrastructure and Ecosystem Development - The establishment of a scientific corpus platform has enabled the collection and management of over 10 PB of high-quality scientific data, facilitating breakthroughs across multiple disciplines [42]. - The integration of AI tools and data governance is crucial for enhancing research efficiency and ensuring data security [39][42]. - The collaborative efforts between the Shanghai Intelligent Research Institute and Fudan University aim to create a comprehensive ecosystem for scientific intelligence, focusing on the integration of computing power, algorithms, and data [35][36].

Core Insights - Artificial intelligence is revolutionizing traditional industries, particularly in the field of new materials, with Beijing leading the way by hosting nearly one-third of the country's "AI + new materials" innovative enterprises [1][11]. Group 1: AI's Impact on Material Science - AI is redefining the research paradigm in material science, significantly enhancing research efficiency [2][3]. - Traditional material research, which often takes 20 to 30 years from discovery to application, is now being accelerated through AI technologies [1][4]. - For instance, a company was able to reduce the research and development cycle for a new electrolyte for electric vehicle batteries from 18 months to approximately 12 months, achieving a one-third reduction in time [3][4]. Group 2: Innovations in Smart Laboratories - Smart laboratories are transforming traditional research environments by automating processes such as material mixing, reaction monitoring, and data analysis, allowing for continuous operation without human intervention [4][5]. - These smart labs can conduct hundreds of automated experiments daily, significantly increasing research output compared to traditional methods [5][6]. - Tasks that previously took 10 to 20 days can now be completed in just 1 to 2 days due to automation and AI integration [6]. Group 3: Successful Applications and Developments - The integration of AI in new materials has led to notable achievements, such as Xiaomi's use of "Titanium Alloy" in their vehicles, which improved structural stability and reduced weight [7][8]. - Other significant developments include the creation of the world's strongest hydrogen-resistant thick plate material by China Steel Research and the design of a new generation of OLED materials without precious metals by the Beijing Academy of Science [8][9]. Group 4: Future Prospects and Strategic Plans - Beijing's strategic plan aims to enhance its "AI + new materials" innovation capabilities by 2027, targeting the development of 15 benchmark new materials products empowered by AI [11]. - The city is positioned as a hub for "AI for Science," leveraging its talent pool, application scenarios, and innovative ecosystem to drive advancements in material science [10][11].