AI Cognition Researcher

To support the AI Cognition Initiative, the full-time remote AI Cognition Researcher will design and conduct empirical experiments focused on understanding large language models as agents, exploring their cognitive mechanisms, and communicating findings to both technical and policy audiences. Key responsibilities Design and implement experiments examining cognitive mechanisms in large language models Investigate aspects such as persona manifestation, preferences, and decision-making processes Communicate research findings clearly to diverse audiences, including policymakers and technical experts Required qualifications Hands-on experience with open-source models and cloud infrastructure Solid understanding of transformer architectures and experimental design Familiarity with modern interpretability techniques, such as activation patching and probing Background in cognitive science, psychology, philosophy of mind, or AI safety is a plus A track record of self-directed research is preferred