Research Scientist (Pre-training) — San Francisco Bay Area
Зарплата не указана
ОткликнутьсяКомпания: ••••••••
Локация: San Francisco Bay Area
Формат: не указан
Занятость: Полная занятость
Грейд: Senior
Principal / Staff Research Scientist – Pre-Training
Frontier AI Research Lab | San Francisco Bay Area | Exceptional Compensation + Equity
We are partnering with a highly ambitious, well-funded frontier AI research lab building next-generation foundation models from the ground up.
We're looking for exceptional senior researchers who have played a meaningful role in training state-of-the-art foundation models at scale.
This is an opportunity to influence fundamental research decisions, shape model architectures and push the boundaries of what is possible in large-scale AI pre-training.
What You'll Work On
• Design and train frontier-scale foundation models, exploring novel architectures, training objectives and methodologies.
• Lead research into scaling laws, data mixtures, optimisation and training efficiency.
• Drive large-scale distributed training experiments across significant GPU compute.
• Translate original research into measurable improvements in model intelligence, reasoning and generalisation.
• Help define the research direction alongside an exceptional team of scientists and engineers.
Who We're Looking For
• Deep, hands-on experience training large-scale foundation models, ideally at a leading AI research lab.
• Significant contributions to pre-training research, model architecture or large-scale training methodologies.
• Strong understanding of transformers, optimisation, scaling laws and distributed training.
• A track record of original research with demonstrable impact on model capabilities.
• The technical depth and ambition to operate at the very frontier of AI research.
Why This Opportunity?
High-impact research. Significant compute. Genuine technical ownership.
An opportunity for researchers who want to do more than improve existing models — they want to help define the next generation.
Interested? Apply directly or reach out for a confidential conversation.