Research Engineer - Pretraining
$150k - $350k • Remote • San Francisco • FullTime
Posted 1d ago
Job Location
San Francisco
Tech Stack
Remote Work Policy
Fully remote
Employment Type
FullTime
Categories
AI Research Engineer
About the job
Prime Intellect is building the open superintelligence stack, providing infrastructure for AI labs. Our platform, Lab, unifies compute, environments, evaluations, secure sandboxes, high-performance training, and deployment for post-training at frontier scale. We are developing open-source models trained end-to-end for long-horizon tasks like autonomous research, alongside the full-stack platform our research team utilizes. We are seeking individuals passionate about building at the intersection of frontier research, real infrastructure, and go-to-market for a nascent category.
Responsibilities
- Build and extend pretraining code, scaling small experiments into full training runs.
- Evaluate model architecture changes for quality and training cost.
- Profile training jobs and improve end-to-end throughput, collaborating on hardware-aware execution.
- Ensure long runs are resilient to failures through reliable checkpointing and recovery.
- Investigate training dynamics and numerical stability to ensure optimization convergence.
- Run controlled ablations to guide larger experiments, ensuring reproducibility.
- Collaborate on training mixtures, including midtraining and long-context experiments.
Requirements
- Hands-on experience training language models, including modifying training code and observing run effects.
- Strong Python skills and experience with PyTorch or a comparable framework, with systems understanding for debugging distributed training.
- Practical understanding of model architectures and optimization, including their impact on convergence.
- Ability to design ablations to answer research questions and interpret results.
- Good judgment for self-investigation and knowing when to seek colleague input.
Benefits
- Cash Compensation Range of $150-350k
- Equity incentives
- Flexible work arrangements (remote or in-person)
- Visa sponsorship
- Relocation assistance
- Quarterly team off-sites
- Hackathons
- Conferences and learning opportunities