Research Engineer / Research Scientist, Human Flourishing

About the job

Responsibilities

• Develop scalable training methods to improve model understanding of human values, support agency, and help achieve long-term goals.
• Lead high-leverage research from open questions to shipped improvements, defining success, running experiments, and integrating into frontier models.
• Build RL environments for difficult-to-verify outcomes and study the generalization of beneficial behavior.
• Create evaluations grounded in human expertise and feedback to reveal capability gaps and unintended incentives.
• Test whether model behavior measures predict better learning, wellbeing, and long-term outcomes to guide training.
• Build new 0→1 capabilities in human domains with increasing impact potential.
• Partner with product teams, engineers, partnerships teams, and domain experts to deploy methods and measure effects.
• Assess AI's impacts on human flourishing at scale.

Requirements

• Deep experience improving frontier models through pretraining or reinforcement learning (RL) / post-training, especially on less verifiable problems.
• Strong LLM skills combined with deep expertise in domains like education or mental health (prior experience not required).
• Unusually strong research record in pretraining or RL / post-training research, especially in a hard-to-verify area.
• Ability to combine strong LLM research skills with deep expertise in education, mental health, or a related field of human benefit.
• Hands-on builder capable of turning ideas into rigorous experiments, writing and debugging code, and working with data and training infrastructure.
• Ability to choose consequential problems and actionable steps.
• Ownership of ambiguous problems and selection of methods based on impact.
• Focus on scalability, generalization, and real-world deployment of capabilities.
• Ability to work across disciplines and understand multidisciplinary bottlenecks.