Researcher, Safety Training, National Security
Core
Research and implement methods for safety training, reinforcement learning, and adversarial robustness to ensure safe behavior in AI models for U.S. government use.
Role type
Senior AI safety researcher (national security)
Builds
Safe, robust AI models deployed in safety-critical national security settings
Domain
AI safety, national security, machine learning
Deliverable
production ML models
Required skills
RLHF, adversarial training, robustness research, deep learning research, model evaluation, failure mode analysis
Preferred skills
experience improving model safety for deployment, collaborative research
Technologies
reinforcement learning, adversarial training frameworks
Responsibilities
Research and implement safety training and adversarial robustness methods; Develop evaluations and identify model failure modes; Work with research, engineering, security, and policy partners to support safe deployment