Research Scientist

• Lead research on loss-of-control and harmful manipulation risks in frontier AI models.
• Design novel evaluation methodologies addressing evaluation awareness, sandbagging, and deception.
• Author safety reports and research publications to advance the field.
• Set research direction and priorities in collaboration with the team.

This role directly contributes to reducing catastrophic risks from advanced AI by identifying and mitigating manipulation, deception, and loss-of-control scenarios, thereby promoting safe and beneficial AI development.

• Expertise in AI safety research, particularly in frontier models and risk evaluation.
• Strong background in machine learning and model evaluation techniques.
• Experience authoring research publications and engaging with the AI safety community.
• Ability to set research direction and collaborate effectively.

Compensation and benefits are competitive and commensurate with experience. Specific details not provided in the posting.

Neo Research is an organization focused on AI safety, conducting cutting-edge research to understand and mitigate risks from frontier AI systems.