Argonne National Laboratory
1 week ago
Postdoctoral Appointee in Federated Learning and Foundation Model Optimization Argonne National Laboratory in United States
Degree Level
Postdoc
Field of study
Computer Science
Funding
The expected hiring range is $72,879–$121,465. The full-time, long-term fixed-term appointment includes comprehensive employee benefits.
Country
United States
University
Argonne National Laboratory
Apply for this position
Free with an ApplyKite account: the application link, the deadline in your tracker and a check of your fit.
Keywords
More like this
Suggested positions
About this position
Argonne National Laboratory is seeking a Postdoctoral Appointee in Federated Learning and Foundation Models to join research in the Mathematics and Computer Science Division in Lemont, Illinois. The work addresses how foundation models learn and how scientific machine learning can be trained when data is large, expensive to process, or cannot leave the site where it was collected.
The research combines mathematical analysis of stochastic-gradient training with the design of efficient, reliable, and private algorithms for distributed and federated learning. The postdoctoral researcher will study optimization and training dynamics for large-scale machine learning, develop and benchmark training algorithms, and investigate efficiency, privacy, robustness, and reproducibility.
The appointee will contribute research software to open-source projects and conduct experiments using high-performance computing resources. Potential research directions include stochastic optimization, heavy-tailed noise, loss landscapes, generalization, differential privacy, parameter-efficient foundation-model adaptation such as LoRA, and multi-node training with technologies including DeepSpeed, PyTorch FSDP, and MPI.
This is a collaborative research environment involving partners at universities and other national laboratories. The appointee will be encouraged to shape an independent research direction, lead publications, present at leading machine-learning and optimization venues, contribute to program reports, collaborate across institutions, and mentor students.
Funding details
The expected hiring range is $72,879–$121,465. The full-time, long-term fixed-term appointment includes comprehensive employee benefits.
What's required
A 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. Applicants must have strong programming skills in Python and PyTorch, including training deep learning models on GPUs, and coursework or research experience in optimization, probability, or machine learning. Clear communication and the ability to work in a team distributed across several institutions are required, along with alignment with Argonne's values of impact, safety, respect, teamwork, and integrity. Desired experience includes stochastic optimization, heavy-tailed noise, loss landscapes, generalization, federated learning, differential privacy, foundation-model fine-tuning or pretraining, parameter-efficient methods such as LoRA, multi-node distributed training with DeepSpeed, PyTorch FSDP, or MPI, and open-source machine-learning software contributions.
How to apply
Apply through the Argonne Careers Workday application page for requisition 423620. Submit the requested application materials through the online portal.
Make your next move.
Prepare your application
Review drafts and confirm requirements with the university.
Make your next move.
Prepare your application
Review drafts and confirm requirements with the university.