AI Safety Research Scientist

Center for AI Safety; San Francisco (CA); On-site; USD 140,000 - 200,000; Full time

Benefits offered by this job

Health Insurance for you and dependets

401K with 4% matching

Unlimited PTO

Lunch and dinner at the office

Annual Professional Development Stip</

Access to top AI safety researchers

Job summary

Center for AI Safety in San Francisco seeks researchers to advance the safety and reliability of frontier AI systems. You will design and run large-scale experiments, train transformers and multimodal models, and turn results into publishable work.

The role emphasizes empirical ML with substantial freedom to pursue high-impact questions, collaboration with academia and industry, and contributions to shared research directions.

Qualifications

• Strong track record of empirical ML research, ideally in AI safety or robustness.
• Ability to read papers and place results in the broader literature.
• Experience designing and debugging large-scale ML experiments.

Responsibilities

• Own research experiments end-to-end, from idea to publication.
• Train and evaluate large transformer models and multimodal systems at scale.
• Build, curate, and maintain datasets and benchmarks for evaluation.
• Collaborate with internal researchers and external partners on shared goals.
• Mentor teammates and improve research infrastructure and reproducibility.

Skills

Empirical ML

Deep learning

PyTorch

Experiment design

Communication

Education

PhD student or researcher in ML

Tools

Compute clusters

Benchmarking frameworks

Job description

Center for AI Safety in San Francisco seeks researchers to advance the safety and reliability of frontier AI systems. You will design and run large-scale experiments, train transformers and multimodal models, and turn results into publishable work.

The role emphasizes empirical ML with substantial freedom to pursue high-impact questions, collaboration with academia and industry, and contributions to shared research directions.