featherlessai - AI Researcher — Training Optimization

Remote (world) - Hybrid6mo ago

In Office WW Artificial Intelligence AI Engineer Reporting Python

Requirements

• Strong background in machine learning research, with emphasis on training dynamics and optimization • machine learning research • training dynamics and optimization • Experience training large neural networks (LLMs, multimodal models, or large sequence models) • large neural networks • Publication experience in ML venues (e.g. NeurIPS, ICML, ICLR, ACL, EMNLP, COLM, arXiv) or equivalent high-quality open research • Solid understanding of: • Optimization theory and practice • Backpropagation, gradient flow, and training stability • Distributed and large-batch training • Proficiency in Python and modern ML frameworks (PyTorch preferred) • Python • Ability to independently design experiments and reason from data • Experience with non-standard architectures (e.g. RNN variants, long-context models, hybrid systems) • non-standard architectures • Experience optimizing training on GPUs at scale (FSDP, ZeRO, custom kernels) • Contributions to open-source ML or research codebases • open-source ML or research codebases • Comfort operating in fast-moving, ambiguous startup environments

Responsibilities

• Design and evaluate training optimization techniques for large models. • Improve training efficiency and stability across long runs and large datasets. • Research and implement methods such as optimizer innovations, mixed-precision training, gradient noise reduction, scaling laws, convergence analysis, training-time regularization, and robustness techniques. • Run large-scale experiments, analyze results, translate findings into actionable improvements for the team's models. • Author or co-author research papers, technical reports, or blog posts to disseminate work in ML communities. • Collaborate closely with infrastructure and inference teams to ensure training decisions are effective when deployed at scale.

Benefits

• Real influence over core model training decisions • core model training decisions • Freedom to pursue and publish novel research • novel research • Direct access to large-scale experiments and real production constraints • A small, senior team that values thinking deeply and shipping thoughtfully • thinking deeply and shipping thoughtfully