研究团队介绍说，磐石·科学文献解析器从底层算法出发，通过构建面向科学语义理解的多模态训 练体系与强化学习机制，在公式、文本、图表等多元素协同解析上实现质的飞跃，为全球科研工作者提 供真正"懂科学"的智能解析引擎。 在磐石·科学文献解析器研发过程中，团队摒弃仅依赖通用视觉语言大模型的思路，转而构建一套 专为科学文献场景量身定制的算法训练范式。其核心在于三大技术支柱：全场景覆盖的科学数据构建、 多模态监督微调策略，以及面向科学文献语义的强化学习优化机制。 在数据层面，系统性采集并构建覆盖手写体、数字排版体与纸质扫描体三大典型科学书写形态的训 练语料，这一"全形态、多学科、高质量"的数据基础，为模型理解科学表达的复杂性提供了坚实支撑。 模型训练阶段采用两阶段优化策略：首先通过多模态有监督微调，使模型初步掌握文本、公式、表 格、插图等异构元素的联合表征能力。在此基础上，引入一种面向科学文献语义的梯度强化学习策略优 化框架，实现模型不仅"看得清"，更能"理解对"。 记者11月1日从中国科学院自动化研究所（自动化所）获悉，该所"AI+科学"研究团队近日正式推出 新一代科学文献解析工具——磐石·科学文献解析器，为全球科研工 ...

Core Insights - The integration of AI and science is emerging as a new trend, providing unprecedented opportunities to address significant technological challenges [1] - The Chinese Academy of Sciences has launched the "Panshi Scientific Foundation Model," a specialized intelligent platform designed for scientific research, capable of deep understanding of various scientific data and advanced literature analysis [1] Group 1 - The "AI + Science" research approach typically employs two methods: fine-tuning general models with specialized data or creating dedicated tools for specific fields, which has led to issues such as data interoperability, insufficient reasoning capabilities, and a closed research ecosystem [1] - To tackle these challenges, a research team from multiple institutes within the Chinese Academy of Sciences has developed the Panshi Scientific Foundation Model, which acts as a cross-disciplinary "intelligent operating system" to manage data and resources effectively [1] Group 2 - Based on the Panshi Scientific Foundation Model, the team has also developed two scientific intelligent agents: the Panshi Literature Compass and the Panshi Tool Scheduling Platform [2] - The Panshi Literature Compass has integrated 170 million scientific documents and real-time open-source scientific information, significantly reducing literature review time from 3-5 days to just 20 minutes [2] - The Panshi Tool Scheduling Platform can autonomously plan and call upon over 300 scientific computing tools, facilitating collaborative orchestration and easy access to these tools [2] - The Panshi Scientific Foundation Model has been applied in various disciplines, notably in high-energy physics, where it enhances the efficiency of particle simulation and reconstruction, aiding in the exploration of fundamental components of matter and universal laws [2]

Core Viewpoint - In 2025, large models are transitioning from laboratories to practical applications across various industries, showcasing significant breakthroughs in multiple scientific fields, including quantum computing, life sciences, materials science, earth sciences, and deep space astronomy [1][2]. Breakthrough Achievements - The world's first AI-based quantum computing neutral atom arrangement algorithm was developed, capable of arranging 2024 quantum bits in 60 milliseconds, overcoming traditional time constraints associated with increasing atom numbers [5]. - The first multi-agent virtual disease scientist system, "OriGene," was introduced, which can automatically discover and validate new treatment targets for cancers, establishing a complete intelligent process from data to verification [6]. - The first single-cell DNA methylation model, scDNAm-GPT, was launched, achieving over 90% accuracy in early detection of various cancers and respiratory diseases using blood samples [7]. - The "Fengdeng" breeding model was established, marking a breakthrough in China's seed industry technology, with over 100 breeding units already testing its applications [8]. - An intelligent scientific discovery system for condensed matter science was released, achieving practical-level standards for new copper-based superconducting materials [9]. - The EarthLink AI Earth scientist system was developed, significantly enhancing research efficiency by 160 times through automated experimental design and data analysis [10]. - An AI tracking system for space debris was created, improving tracking accuracy by 70% compared to traditional methods [11]. - The ChemBOMAS multi-agent system optimized chemical reactions, reducing precious metal catalyst usage by 90% and increasing yield from 20% to 96% [12]. - The RNA virus language model, Viracle, was introduced, providing high-precision predictions for RNA viruses, with over 95% accuracy for human viruses [13]. - A 3D aircraft generation agent was developed, streamlining the aircraft design process through natural language interaction and rapid concept generation [14]. Future Directions - The Shanghai AI Laboratory plans to continue advancing the integration of specialized technologies to address key scientific discovery challenges, leveraging the "Shusheng" scientific discovery platform for innovation [15].