Foundation Model Research Intern

Invitation only - train large-scale foundation models end to end. You'll write PyTorch, run distributed training jobs, build and debug data pipelines, run scaling-law ablations, build internal (and possibly future public) benchmarks, and turn research ideas into production model checkpoints. Expect to care deeply about throughput, loss curves, and why a run went sideways at 3 a.m.

Full-time available