Research Intern AI Safety

You will design and implement AI safety guardrails for public safety, including dynamic benchmarking, red-team simulations, harm detection, and automated evaluation. You will document and publish research, collaborate on experiment design and technology integration, identify applied research opportunities, and optimize model research, training, and inference.

Responsibilities

• Design and implement AI safety guardrails for public safety
• Develop dynamic benchmarking, red-team simulations, harm detection, and automated evaluation
• Create internal technical documentation and write a scientific manuscript
• Collaborate with product managers on experiment-design requirements
• Select, evaluate, and integrate AI technologies with researchers
• Identify applied research opportunities from prototypes and user feedback
• Optimize research, model training, and inference processes

Requirements

• AI research experience
• Multiple publications at major conferences, including NeurIPS, ICML, or ICLR
• Applied AI safety research
• Scientific communication
• Multidisciplinary collaboration