Jasmine Li

Jasmine is a Research Scientist on SaferAI’s Frontier AI Risk Management and team, developing evaluations and risk assessments for frontier models. Before joining SaferAI, she was a MATS Fellow with Alex Turner working on training mitigations for evaluation gaming, built agent security infrastructure and benchmarks at Gray Swan AI, and researched LLM honesty with the Center for AI Safety. She studies Computer Science at Cornell University.