Lawrence Livermore National Laboratory (LLNL) is seeking a postdoctoral researcher to advance fundamental R&D at the intersection of reduced-order modeling, foundation models for computational sciences, and statistical/machine learning methods to develop scalable models and methods for materials science, fusion, and additive manufacturing.
Requirements
- Demonstrated expertise in spatial discretization methods (finite element, finite volume, and finite difference methods).
- Strong background in numerical time integration techniques. Demonstrated experience with numerical analysis, including stability and convergence.
- Experience building and evaluating modern ML models using PyTorch, TensorFlow, and/or JAX.
- Significant software development experience with the Python scientific software stack; demonstrated experience following modern software engineering practices (e.g. testing, version control, reproducibility).
- Significant experience building and evaluating reduced order models using Proper Orthogonal Decomposition, Dynamic mode decomposition, and/or hyper-reduction.
- Significant experience using the libROM software library or equivalent.
- Demonstrated research productivity, as documented by publications, reports, presentations, and/or open-source software, in relevant venues (NeurIPS, ICML, JCP, CMAME, Science, IJNME, SISC, Nature etc.).
Responsibilities
- Research & prototype hybrid reduced order model/machine learning methods (DD-FEM, LaSDI, gappy Autoencoder, proper orthogonal decomposition, and neural operators) tightly coupled to governing equations.
- Design discretizations & integrators: derive/implement stable and accurate spatial discretizations (e.g., finite element, finite volume, finite difference methods) and time integrators (e.g., explicit/implicit Runge–Kutta, multistep/BDF).
- Guarantee physics & reliability: enforce conservation/stability, perform error and sensitivity analysis, and lead UQ, and explainability for hybrid models.
- Integrate with HPC codes and experimental/simulation workflows.
- Publish results, contribute proposals, and collaborate across disciplines.
- Perform other duties as assigned.
Other
- Must be eligible to access the Laboratory in compliance with Section 3112 of the National Defense Authorization Act (NDAA).
- Ph.D. in Computational Science, Applied Mathematics, Engineering, Statistics, or a related field.
- Demonstrated ability and desire to obtain substantial domain knowledge in fields of application to enable effective communication with subject matter experts, and to identify novel, impactful applications of machine learning.
- This is a Postdoctoral appointment with the possibility of extension to a maximum of three years, open to those who have been awarded a PhD at time of hire date.
- External applicant(s) selected for this position must pass a post-offer, pre-employment drug test.
