来源：联想创投 近日，联想创投被投企业「深度原理 Deep Principle」完成超亿元人民币A轮融资。本轮由戈壁创投管理的阿里巴巴创业者基金大湾区基金(简称AEF大湾 区基金)与蚂蚁集团共同领投，现有股东联想创投、Taihill Venture 超额加注，BV百度风投继续加注，多家机构参与。 本轮融资将主要用于三大方向：加速 Agentic AI for Materials Discovery 材料发现智能体 Agent Mira™的研发与升级；推进 L4 高通量自主实验室 AI Materials Factory™与其研发管线的建设与布局；深化与国际和国内头部客户的合作，巩固技术落地领先优势。 顶尖团队驱动模型持续突破 联想集团副总裁、联想创投管理合伙人王光熙表示，联想创投本轮加注深度原理，核心是持续看好 AI for Science 引领的产业变革机遇，更认可深度原理 的技术硬实力与商业化闭环能力。团队以扩散模型与大语言模型的前沿突破为根基，将材料研发从"试错式实验"推向"智能化设计"，凭借显著技术领先 性，成立仅一年便斩获超千万订单，展现了团队将顶尖学术成果向产业解决方案转化的超强效率。其提出的 E ...

Core Insights - The article discusses the strategic financing of "Deep Principle," a leading company in the AI for Science sector, which has recently completed a significant funding round to enhance its technological capabilities and market presence [3][4]. Group 1: Financing and Investment - Jinqiu Fund participated in a strategic Pre-A round financing for "Deep Principle," amounting to over 100 million RMB [3]. - The latest A round financing was led by Alibaba's Entrepreneur Fund and Ant Group, with participation from existing investors such as Lenovo Ventures and Baidu Ventures [4]. Group 2: Technological Advancements - "Deep Principle" has developed cutting-edge diffusion models, achieving significant breakthroughs in AI for materials discovery, with their OA-ReactDiff model completing transition state structure predictions in just 6 seconds [7]. - The company has introduced the React-OT model, which reduces prediction time to 0.4 seconds and improves accuracy by over 25% compared to traditional methods [7]. Group 3: Product Development and Applications - The company has launched the Agent Mira platform, which integrates various algorithms for molecular design, chemical reaction prediction, and materials optimization, making AI a routine part of industrial processes [10][12]. - "Deep Principle" has secured over 10 million RMB in commercial orders within its first year, collaborating with industry leaders like L'Oréal and focusing on applications in supramolecular materials and fine chemicals [12]. Group 4: Future Directions - The company is building an L4 High-Throughput Autonomous Lab, known as AI Materials Factory, to streamline the materials discovery process and enhance the integration of AI models with experimental validation [14][15]. - The recent financing will serve as a new starting point for "Deep Principle" to deepen the integration of technological innovation with industry needs, contributing to the advancement of global materials science [16].

Core Viewpoint - Deep Principle, a pioneer in AI for Science, has completed over 100 million RMB in Series A financing, led by AEF Bay Area Fund and Ant Group, with participation from existing shareholders and multiple institutions [2] Financing Purpose - The financing will primarily be used to accelerate the development and upgrade of the Agent Mira™ for materials discovery, advance the construction of the L4 High-Throughput Autonomous Lab AI Materials Factory™, and deepen collaborations with leading domestic and international clients to consolidate technological advantages [3] Technological Leadership - Deep Principle was founded by a team with a background from MIT, specializing in AI for Science. The company has developed diffusion models that have been published in top journals, achieving significant breakthroughs in predicting transition state structures in a fraction of the time compared to traditional quantum chemistry methods [4] - The company introduced the OA-ReactDiff model in 2023, which can predict transition states in 6 seconds on a single GPU, and plans to release an upgraded model, React-OT, in early 2025, which will reduce prediction time to 0.4 seconds and improve accuracy by over 25% [4] Product Upgrade - The company has developed six algorithm modules integrated into the ReactiveAI platform, which has recently been upgraded to the Agent Mira for materials discovery. This platform allows for molecular structure design, chemical reaction prediction, and materials formulation optimization through natural language commands [7] Commercialization Progress - Deep Principle has achieved substantial commercial progress, securing over 10 million RMB in commercial orders within its first year. Collaborations include building an AI supramolecular materials platform with Shanhai Innovation and working with L'Oréal to enhance formulation performance through AI [8][9] AI Materials Factory - The company is establishing the L4 High-Throughput Autonomous Lab AI Materials Factory, which integrates the ECML research paradigm to create a closed loop of AI model prediction, computational support, and experimental validation, facilitating the application of its technology in key strategic areas [11] Future Outlook - With this round of financing, Deep Principle aims to deepen the integration of technological innovation with industry needs, continuously injecting AI-driven momentum into the development of global materials science [12]