AI Safety & Evals Scientist/Engineer

Apolloresearch; City of London; Hybrid; GBP 100,000 - 200,000; Full time

Job summary

A leading AI research firm in London seeks Research Scientists and Engineers to evaluate and mitigate risks in AI models. Ideal candidates will collaborate with frontier labs and contribute to innovative projects related to scheming. The role requires strong experience in empirical research, LLM steering, and software engineering. Flexible hours, unlimited vacation, and a competitive salary of £100k to £200k are offered.

Benefits

Flexible work hours

Unlimited vacation

Paid work trips

Professional development budget

Qualifications

• Self-taught candidates welcomed.
• Experience in empirical research related to scheming and AI.
• Strong scientific writing and communication skills.
• Comprehensive experience in LLM steering.
• Software engineering skills required.

Responsibilities

• Work on safety evaluations for frontier models.
• Design and evaluate AI control protocols.
• Automate evaluation pipelines.
• Collaborate with frontier labs.

Skills

Empirical research

Scientific writing

LLM steering

Software engineering

Tools

Python

Inspect

Job description

A leading AI research firm in London seeks Research Scientists and Engineers to evaluate and mitigate risks in AI models. Ideal candidates will collaborate with frontier labs and contribute to innovative projects related to scheming. The role requires strong experience in empirical research, LLM steering, and software engineering. Flexible hours, unlimited vacation, and a competitive salary of £100k to £200k are offered.