Staff / Principal Research Engineer, AI Safety Mitigations
$224k - $336k • Cambridge, MA USA; London, UK; San Francisco, CA USA • Full-time
Posted 15d ago
Job Location
Cambridge, MA USA; London, UK; San Francisco, CA USA
Remote Work Policy
On-site
Employment Type
Full-time
Categories
AI Research Engineer
About the job
Lila Sciences is building Scientific Superintelligence™ to accelerate discovery across medicine, materials, and energy. We are seeking a talented individual contributor to join our AI safety team, focusing on identifying and mitigating risks from scientific superintelligence. This role involves developing and implementing a bespoke safety strategy tailored to Lila's specific goals and deployment strategies. You will work with domain scientists, risk researchers, evaluators, and engineering teams to translate risk findings and policy requirements into effective interventions, spanning research and implementation. The work includes developing approaches, testing their limits, and integrating reliable safeguards into our systems.
Responsibilities
- Translate scientific risk assessments, dual-use policy boundaries, and evaluation findings into concrete mitigation designs and testable requirements.
- Develop and implement safeguards suited to the problem, potentially spanning model training, classifiers, monitoring, access controls, and runtime interventions.
- Build experiments to measure risk reduction, robustness, and effects on legitimate scientific use, alongside practical performance constraints.
- Use adversarial findings and observed failures to improve safeguards, collaborating with automated-safety and scientific evaluation researchers.
- Integrate mitigations into model development, release evaluation, and large-scale serving systems with appropriate instrumentation and reproducible testing.
- Investigate regressions, distribution shifts, and emerging misuse patterns to guide further development.
Requirements
- Hands-on experience developing and validating safety interventions for frontier models within training, post-training, or large-scale inference systems.
- Depth in at least one relevant area: safety post-training, model-level classifiers, safeguards integrated into inference, or abuse monitoring and enforcement.
- Experience working directly with model-development or safeguards teams to investigate failures and implement improvements.
- Strong research-engineering skills with the ability to move from ambiguous problems to working, maintainable implementations.
- Sound experimental judgment, including assessing robustness, interpreting imperfect metrics, and reasoning about competing objectives.
- Ability to collaborate across research and engineering teams, explain technical trade-offs, and take ownership of complex individual contributions.
Benefits
- Competitive base compensation with bonus potential
- Generous early-stage equity
- Comprehensive U.S. benefits program including medical, dental, and vision coverage
- Employer-paid life and disability insurance
- Flexible time off with generous company wide holidays
- Paid parental leave
- Educational assistance program
- Commuter benefits including bike share memberships
- Company subsidized lunch program
- International benefits tailored to region