Explainable SI - Postdoctoral Researcher

LLNL • Livermore, CA

23h • Hybrid

About The Position

Lawrence Livermore National Laboratory is hiring a Postdoctoral Researcher in Explainable SI to conduct fundamental research on what modern SI models learn and how that knowledge is represented internally. The role focuses on recovering human-meaningful structure from learned representations through sparse decompositions of activations, concept discovery, mechanistic analysis, and causal intervention. It also develops interactive interfaces and evaluation methods that allow domain experts to interrogate model internals and explanations. Applications include multimodal SciML models and deep surrogates for simulations. The position is in the Machine Intelligence Group in the Center for Applied Scientific Computing (CASC) Division within the LLNL Computing Directorate.

The researcher will develop and evaluate interpretability methods, create human-in-the-loop workflows for domain experts, and establish rigorous evaluation of interpretability claims. They will conduct independent machine learning research and collaborate with project scientists, engineers, and multidisciplinary partners on complex scientific and national security applications. Duties include proposing and applying analysis methodologies, collecting and analyzing data, documenting results, publishing research, contributing to grant proposals, and pursuing complementary independent research interests.

This is a temporary, mentored postdoctoral appointment with the possibility of extension to a maximum of three years. Applicants must have received their Ph.D. or equivalent doctorate within five years of the anticipated start date. The role offers a hybrid schedule, with the flexibility to work from home one or more days per week.

Requirements

• Recent Ph.D. in Computer Science, Machine Learning, Applied Mathematics, Statistics, Human-Computer Interaction, or a related field.
• In-depth knowledge of explainable SI and related topics, demonstrated through relevant experience and publications.
• Research experience in explainable or interpretable SI, representation learning, mechanistic interpretability, concept-based explanation, or visual analytics for machine learning.
• Experience developing and applying medium- to large-scale deep learning methods using modern libraries such as PyTorch or JAX.
• Demonstrated research productivity documented by publications, reports, presentations, and/or open-source software in relevant venues.
• Experience with scientific programming in the Python ecosystem.
• Ability to gain substantial domain knowledge in application fields and communicate effectively with subject matter experts.
• Must have received a Ph.D. or equivalent doctorate within five years of the anticipated start date.

Nice To Haves

• Experience with sparse autoencoders, transcoders, or related feature-learning methods applied to activations of large pretrained models.
• Experience analyzing or intervening on internal representations of trained models, such as probing for encoded properties, steering or editing activations, or attributing outputs to internal components.
• Experience connecting interpretability to uncertainty quantification, robustness, calibration, or SI safety and assurance evaluation.
• Experience with high-performance computing, GPU programming, parallel programming, cloud computing, related methods, or numerical simulations of complex workflows.
• Demonstrated technical leadership in machine learning-related fields, such as mentorship or managing teams.
• Experience or interest in scientific applications such as material science or climate science.

Responsibilities

• Develop and evaluate methods for interpreting internal representations of deep models, including sparse decompositions of activations, concept extraction, and representation steering.
• Design human-in-the-loop workflows that allow domain experts to explore, validate, and correct discovered concepts, and evaluate those workflows with real users.
• Establish rigorous evaluation methodology for interpretability claims, including faithfulness, stability, and causal grounding.
• Research, design, implement, and apply advanced machine learning methods across multiple applications in a collaborative scientific environment.
• Conduct cutting-edge machine learning research effectively and independently.
• Participate with project scientists and engineers in defining, planning, and formulating experimental, modeling, and simulation efforts for complex national security problems.
• Propose and implement advanced analysis methodologies, collect and analyze data, and document results in technical reports and peer-reviewed publications.
• Contribute to grant proposals and collaborate with multidisciplinary teams, including academic and industrial partners, to accomplish research goals.
• Pursue independent but complementary research interests and interact with scientists inside and outside the Laboratory.
• Perform other duties as assigned.

Benefits

• Flexible benefits package
• 401(k)
• Relocation assistance
• Education reimbursement program
• Flexible schedules, depending on project needs

What This Job Offers

• Job Type: Full-time
• Career Level: Intern
• Education Level: Ph.D. or professional degree