AI4Science研发实习生

职位独特卖点 (USP):

1. 我们是博世人工智能研究中心（中国），聚焦AI与工程系统深度融合（AI4Engineering），以全球规模的真实汽车应用为驱动，桥接前沿AI与复杂工程系统

2. 本职位独特地结合了数值与人工智能方法（如有限元方法、神经算子、扩散模型等），并在与顶尖高校合作伙伴的紧密合作中，推动基于PyTorch的建模与系统辨识方向的创新研究

3. 团队成员90%以上拥有博士学位，来自全球顶尖院校，兼具深厚的科研探索能力与在复杂工业系统中落地AI解决方案的实战经验

职位内容:

1. 面向真实场景应用，开展AI4Engineering相关研究，重点聚焦于科学机器学习、基于PyTorch的多物理场仿真建模以及大规模物理引导反问题等方向

2. 设计与开发先进的AI算法与模型，利用真实电子数据进行电路设计与优化

3. 与顶尖高校合作伙伴紧密协作，打通"理论建模—算法开发—原型验证—量产落地"的全流程，并积极推动顶级期刊会议论文发表及专利技术成果输出

4. 跟踪AI4Engineering与科学机器学习等前沿方向，结合工业需求识别高价值问题，并推动创新技术在实际场景中的规模化应用

1. 具备扎实的编程能力，熟练使用PyTorch等深度学习框架；有科学机器学习经验者高度优先

2. 对求解大规模物理引导反问题的数值与人工智能方法（如有限元方法、扩散模型、神经算子、基于采样的方法等）有深入理解或浓厚兴趣

3. 英语听说读写流利，具备强大的跨团队、跨文化科研与工程协作能力

4. 拥有计算机、人工智能、应用数学、计算物理或相关专业博士学位

5. 在相关领域顶级期刊或会议发表论文者优先

Unique Selling Points:

1. We are the Bosch Center for Artificial Intelligence (China), focusing on AI4Engineering—bridging cutting-edge AI with complex engineering systems, driven by real-world automotive applications at global scale

2. This role uniquely combines numerical and AI methods (e.g., finite-element method, neural operators, diffusion models), advancing innovation in PyTorch-based modeling and system identification in close collaboration with top academic partners

3. Over 90% of our team members hold PhDs from leading global institutions, combining deep research expertise with hands-on experience in deploying AI solutions in complex industrial systems

Responsibilities:

1. Conduct AI4Engineering research for real-world applications, focusing on scientific machine learning, PyTorch-based modeling for Multiphysics simulation, and large-scale physics-based inverse problems

2. Develop advanced AI algorithms and models using real electronic data for circuit design and optimization

3. Collaborate closely with top academic partners to bridge the full pipeline from theoretical modeling, algorithm development, prototyping to production deployment, and contribute to top-tier publications and patents

4. Stay at the forefront of AI4Engineering and Scientific Machine Learning, identify high-impact opportunities aligned with industrial needs, and drive scalable deployment of innovations

Requirements:

1. Strong programming skills, proficiency in deep learning frameworks (e.g., PyTorch); experience in scientific machine learning is highly preferred

2. Solid understanding of or strong interest in numerical and AI methods for solving large-scale physics-based inverse problems (e.g., finite-element method, diffusion models, neural operators, sampling-based method)

3. Fluent in English with strong cross-team and cross-cultural collaboration skills

4. PhD in computer science, AI, applied mathematics, computational physics, or related fields

5. Publications in top-tier journals or conferences are a plus

Qualifications:

1. PhD candidates in computer science, applied mathematics, or machine learning are preferred. Candidates of Bachelor's or Master's in Science with exceptional mathematical and programming skills can also be considered.

2. Proficiency in at least one programming language (preferably in Python) and experience in GPU computing.

3. Strong background in theory of physics simulation, inverse problems, Bayesian inference, and familiarity with various architectures in deep learning.

4. Capability of independent work, passion in AI research, and good communication skills.