Overview In this role you advance post-training and reinforcement learning methods for scientific foundation models, aiming to improve usefulness, reliability, and scientific impact. You will scale algorithms and pipelines on leadership-class supercomputers, collaborating with domain scientists to solve high-impact problems. The position sits in a multidisciplinary AI group at Argonne, contributing foundational ML research and real-world scientific outcomes. You'll publish, collaborate with national labs and academia, and help drive the AI for science mission.
Compensation / Benefits- Hybrid remote work - Mostly onsite
- comprehensive benefits package
Responsibilities- Develop, scale, and optimize post-training methods for scientific foundation models
- Advance techniques to improve performance, controllability, reliability, and scientific utility of AI models
- Design and evaluate RL and post-training pipelines for large-scale scientific environments
- Develop and optimize workflows for training and post-training on supercomputers and AI-oriented architectures
- Collaborate with computational scientists and domain researchers on challenging scientific problems
- Address algorithmic, systems, and data challenges in large-scale training and post-training
- Conduct original research and communicate findings through publications, talks, and software
- Engage with national labs, universities, industry, and supercomputing centers on AI for science initiatives
- Foster a team culture valuing scientific excellence, collaboration, and inclusive growth
Key requirements- Bachelor's degree with 5+ years of experience, or a Master's with 3+ years, or a PhD, in a related field
- Advanced knowledge in machine learning, reinforcement learning, large-scale model training, post-training, optimization, data mining, or statistics
- Strong background in mathematical optimization, linear algebra, or numerical methods
- Programming experience in Python, C, or C++
- Experience with ML frameworks such as PyTorch or JAX
- Experience with large-scale training, distributed learning systems, or post-training workflows
- Experience with software development practices for computational science and ML systems
- Ability to work effectively in interdisciplinary teams
- Effective written and verbal communication skills
- Ability to model Argonne's core values of impact, safety, respect, integrity, and teamwork
- collaborative mindset
- strong communication
- teamwork in interdisciplinary settings
- reinforcement learning
- post-training methods
- fundamental ML and DL techniques
