Research Scientist: AI Self-Improvement & Evaluation

Anthropic; San Francisco (CA); Hybrid; USD 350,000 - 850,000; Full time

Benefits offered by this job

Competitive compensation

Equity donation matching

Generous vacation and parental leave

Flexible working hours

Lovely office space

Job summary

Anthropic in San Francisco seeks a Research Scientist to measure recursive-self-improvement in large models. You will design evaluations, build models of capability growth, and interpret results to guide research direction.

Senior candidates will combine hands-on work with strategic planning, collaborating across pretraining, RL, and policy teams to advance safe and reliable AI systems.

Qualifications

• Hands-on research on large language models: pretraining, fine-tuning, RL, evals, or agent systems.
• Strong quantitative instincts and ability to reason with data.
• Experience in forecasting AI trends or scenarios.

Responsibilities

• Identify signals that track AI R&D acceleration and design measurements.
• Build quantitative models of capability growth and self-improvement dynamics.
• Run experiments and evals to test hypotheses about automation and capability.
• Make opinionated research bets and own the outcome.
• Write graded assessments of measurements for internal decisions and public reporting.
• Collaborate with pretraining, RL, economic research, and policy teams.

Skills

Large models research

Quantitative modeling

Forecasting AI trends

Experiment design

Education

Bachelor's degree

Job description

Anthropic in San Francisco seeks a Research Scientist to measure recursive-self-improvement in large models. You will design evaluations, build models of capability growth, and interpret results to guide research direction.

Senior candidates will combine hands-on work with strategic planning, collaborating across pretraining, RL, and policy teams to advance safe and reliable AI systems.