Scientist/Sr. Scientist, AI Safety
Lila Sciences is seeking a Scientist or Senior Scientist specializing in AI Safety to join our high-agency AI safety team. This team collaborates across all core departments, including science, model training, and lab integration, to address potential risks associated with scientific superintelligence. The primary objective is to develop and implement a tailored safety strategy aligned with Lila's specific goals and deployment plans.
In this role, you will be responsible for designing evaluations to identify both known and novel scientific risks arising from advanced scientific models integrated with automated physical laboratories. You will also develop initial safeguards, such as machine learning models, to detect and prevent unsafe behaviors in scientific AI models and physical lab outputs. Additionally, you will assess a range of model capabilities across scientific and non-scientific domains, including persuasion and deception, to inform Lila's comprehensive safety strategy. Your work will contribute to high-quality research efforts focused on evaluating and mitigating scientific capabilities as needed.
Candidates should possess a bachelor's degree in a technical field such as computer science, engineering, machine learning, mathematics, physics, or statistics, or have equivalent experience. Strong programming skills in Python and experience with machine learning frameworks, including tools like Inspect for large-scale evaluation and scaffolded testing, are essential. Experience in building evaluations or conducting red-teaming exercises for CBRN or cyber risks, or for frontier model capabilities in general, is required. Additionally, experience in designing or implementing AI safety frameworks for frontier AI companies and the ability to effectively communicate complex technical concepts to non-expert audiences are necessary.
Preferred qualifications include a master's or PhD in a field relevant to safety evaluations of AI models in scientific domains or a related technical field. Publications in AI safety, evaluations, or model behavior in top machine learning or AI conferences such as NeurIPS, ICML, ICLR, or ACL, or contributions to model release system cards, are advantageous. Experience researching risks from novel science areas like biosecurity or computational biology, or working with specialized scientific tools such as large-scale foundation models for science, is also beneficial.
This position offers a competitive base salary ranging from $228,000 to $358,000 USD, along with bonus potential and generous early-stage equity. Full-time U.S. employees receive a comprehensive benefits package, including medical, dental, and vision coverage; employer-paid life and disability insurance; flexible time off with generous company-wide holidays; paid parental leave; an educational assistance program; commuter benefits, including bike share memberships for office-based employees; and a company-subsidized lunch program. For employees outside the U.S., a comprehensive benefits program tailored to the region is provided.