Research Lead: AI Safety & Pre-Training at Scale

FAR.AI, a Berkeley-based AI safety research institute, seeks a Research Lead to define and own a research program on pre-training safety and capability control. You will lead a growing team, mentor researchers, and push for empirical, scalable safety methods from data filtering to model governance.

You will collaborate with red-teaming partners and governments, drive research roadmaps, and publish results to advance the field while maintaining a strong hands-on coding and experimentation