Postdoctoral Appointee - Federated Learning and Foundation Model

Argonne National Laboratory • Lemont, IL

4d

About The Position

Argonne National Laboratory is seeking a postdoctoral appointee to research methods for training foundation models on scientific data, particularly where scale, cost, and restrictions on moving data present challenges. The research spans the mathematical analysis of stochastic-gradient training and the design of efficient, reliable, and private algorithms for distributed and federated learning. The appointee will be part of the Mathematics and Computer Science Division, collaborate with universities and other national laboratories, and have access to powerful computing systems. Appointees are encouraged to shape their research direction, publish at leading machine learning and optimization venues, and lead papers. They will also help mentor students and develop research independence and professional skills.

Requirements

• Ph.D. completed within the last 0–5 years in a relevant field, with a publication record in optimization for machine learning, deep learning theory, or distributed or federated learning.
• Strong programming skills in Python and PyTorch, including training deep learning models on GPUs.
• Coursework or research experience in optimization, probability, or machine learning.
• Clear communication skills and ability to work in a team spread across several institutions.
• Ability to model Argonne's core values of impact, safety, respect, teamwork, and integrity.

Nice To Haves

• Background in stochastic optimization, heavy-tailed noise, or analysis of loss landscapes and generalization.
• Experience with federated learning, differential privacy, or other privacy-preserving methods.
• Experience fine-tuning or pretraining foundation models, including parameter-efficient methods such as LoRA (low-rank adaptation).
• Experience with multi-node distributed training, such as DeepSpeed, PyTorch FSDP, or MPI.
• Contributions to open-source machine learning software.

Responsibilities

• Conduct research on optimization and training dynamics for large-scale machine learning, combining analysis with careful experiments.
• Design, implement, and benchmark training algorithms for distributed and federated settings, considering efficiency, privacy, and robustness.
• Contribute research software to open-source projects and keep experiments reproducible.
• Publish and present results at leading conferences and journals, and contribute to program reports.
• Work with collaborators across institutions.
• Help mentor students.
• Develop research independence and professional skills, including communication, networking, and leadership.

What This Job Offers

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