SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 180,000 - 298,000 / annual
Lila Sciences is building Scientific Superintelligence to solve major scientific challenges through autonomous AI systems. They are assembling a talent-dense AI safety team to identify and mitigate risks from scientific superintelligence while preserving core discovery capabilities.
In this role, you will develop research methods and tools for safety assessment and safeguards as scientific AI systems become more capable and autonomous. Your initial focus will be on misuse prevention: building flexible, automated red-teaming approaches that test safety-relevant scientific capabilities and the robustness of safeguards.
Key responsibilities include:
- Develop and run capability evaluations of scientific models, agents, and tool-mediated workflows (including simulated lab environments), collaborating with domain experts to ensure tests are scientifically meaningful
- Create methods or systems that explore diverse failure modes, adapt tests as evidence emerges, and support systematic investigation beyond fixed benchmarks
- Build reproducible experiments and evaluation tools that capture relevant behavior, outcomes, and evidence needed to interpret findings
- Measure the quality and coverage of automated assessments, including failure cases in the evaluators themselves
- Develop broader safety automation across threat model management, evidence triage, regression testing, and monitoring workflows
- Collaborate with risk researchers and mitigation engineers to translate findings into improved assessments and safeguards, and verify effectiveness
The role welcomes both Research Scientists and Research Engineers; responsibilities can span or evolve between the two profiles.
Qualifications:
- Strong empirical judgment, practical programming ability, and experience working with modern AI systems
- Ability to investigate ambiguous failures, design informative comparisons, and communicate limitations clearly
- Interest in scientific misuse risks and ability to collaborate productively with domain specialists
- For Research Scientist profile: evidence of selecting valuable research questions, developing or adapting methods, and designing experiments that yield defensible conclusions; ability to prototype and test ideas
- For Research Engineer profile: evidence of building reliable research software, evaluation infrastructure, or agent systems that enable reproducible experiments and efficient iteration; ability to contribute to experimental design and interpretation
Bonus qualifications:
- Automated red teaming, adversarial testing, or evaluation-method research
- Scientific AI, tool-using agents, laboratory automation, or other complex environments
- Experience with safety monitoring, model-based evaluators, or scalable experimental infrastructure
