Research Engineer at Aisafety

We are the Center for AI Safety, a non-profit in San Francisco focused on technical research that reduces high-consequence risks from advanced AI systems. We are hiring a senior research engineer to join our on-site team and help expand the field through rigorous experiments and community-building work.

Responsibilities

• Design and run experiments that fine-tune large-scale transformers and evaluate them across varied data domains such as HarmBench.
• Create new datasets and benchmarks that test model robustness and safety properties.
• Assess model behavior inside sequential decision-making environments and other complex settings.
• Launch ML competitions and workshops that strengthen the AI safety research community.
• Collaborate with internal researchers and external academics on projects covering transparency, robustness, honest AI, and related topics.
• Prepare manuscripts and submit work to leading machine-learning conferences.
• Help organize events and initiatives that connect the global research community with AI safety priorities.

Must-haves

• Degree in computer science or a closely related machine-learning field.
• Co-author on an NLP paper accepted to a top-tier conference.
• Strong alignment with our mission to reduce societal-scale AI risks.
• Ability to read an ML paper, extract its core contribution, and place it in the broader literature.
• Hands-on experience with PyTorch and Hugging Face libraries.
• Experience launching and training distributed ML jobs that scale to hundreds or thousands of GPUs.
• Clear, prompt communication with both technical and non-technical teammates.
• Fluency in English.

Nice-to-haves

• First-author publication at a top conference.
• Well-formed ideas about the most tractable technical problems in AI safety.
• Track record of organizing workshops, competitions, or other community events at major ML conferences.

Benefits

• Health, dental, and vision insurance for you and your dependents, with 100 % coverage for employees.
• 401(k) plan with 4 % employer match.
• Commuter benefits.
• Lunch and dinner provided at the office.
• Personalized ergonomic workstation setup.
• Direct collaboration with leading researchers in technical and conceptual AI safety.