Frontier AI Safety Research Scientist (On-site SF)

TTN Talent; San Francisco (CA); On-site; USD 405,000 - 495,000; Full time

Benefits offered by this job

Equity

Job summary

TTN Talent is seeking a hands-on AI safety researcher to help pioneer a new evaluation direction for frontier models. You will design benchmarks and environments to measure advanced model capabilities and misaligned behavior, working closely with engineers to translate research into reproducible evaluations used by top AI organisations.

The role is based on-site in San Francisco, five days per week, with salary up to $450,000 plus equity.

Qualifications

• Recent hands-on research experience in AI safety and alignment for frontier models.
• Contributions to recognised benchmarks, published research or public evaluations.
• Hands-on experience with reinforcement learning, post-training or frontier-model evaluations.
• Experience with evaluating areas like reward hacking, deception, sabotage or AI R&D capability.
• Ability to design rigorous benchmarks and environments and work with engineers to create reproducible evaluations.

Responsibilities

• Design benchmarks and environments that measure advanced model capabilities and misaligned behaviour.
• Collaborate with engineers to turn research into evaluations used by leading AI organisations.
• Own the new safety evaluation direction from concept to usable assessments, on-site in San Francisco.

Skills

AI safety research

Frontier models

Reinforcement learning

Benchmark design

Evaluation development

Job description

TTN Talent is seeking a hands-on AI safety researcher to help pioneer a new evaluation direction for frontier models. You will design benchmarks and environments to measure advanced model capabilities and misaligned behavior, working closely with engineers to translate research into reproducible evaluations used by top AI organisations.

The role is based on-site in San Francisco, five days per week, with salary up to $450,000 plus equity.