Mathematical Scientist for AI Safety Research

You will explore AI architectures, theoretical frameworks, and safety mechanisms. You will formalize requirements for safe AI systems, develop probabilistic safety guarantees, analyze theoretical research, help prioritize experiments, contribute to scientific papers, and share AI safety knowledge with researchers and engineers.

Responsibilities

• Explore AI system architectures, theoretical frameworks, and safety mechanisms
• Analyze and refine formal requirements for safe AI systems
• Develop probabilistic safety guarantees under finite data and compute constraints
• Contribute to theoretical research and scientific papers
• Study frontier-model risks and relevant machine learning research
• Guide and prioritize safety experiments
• Help design machine learning algorithms and experiments
• Travel to conferences with team members when needed

Requirements

• PhD in machine learning, computer science, mathematics, statistics, or an adjacent field
• Research track record in formalization and technical analysis
• Interest or experience in technical AI safety
• Understanding of probabilistic machine learning foundations for neural networks and language models
• Written and verbal communication skills

Benefits

• Comprehensive health benefits including a mental health and wellness management account
• 20 days of vacation per year upon start
• 4% employer retirement savings contribution with no required employee match
• Additional compensation totaling 8% of salary for retirement savings or bonuses