I am an independent researcher working at the intersection of AI behavioral evaluation, post-training, and comparative behavioral science.

My current project, COOP-Transfer, investigates whether training across different human-agent contingencies can produce cooperative behavior that generalizes appropriately to novel contexts. The work focuses on questions including evidence-seeking behavior, stimulus control, behavioral generalization, calibration, over-deference, and context-sensitive cooperation in LLM agents.

My background is unusual for AI research. Before moving into technical systems work, I spent years in applied animal behavior, including clinical behavior modification and shelter behavior-program design. I later moved into IT and cybersecurity, eventually managing service delivery and technical operations for an MSP. That combination led me to approach LLM behavior as an experimental behavioral system: define observable behavior, manipulate contingencies, test generalization, and distinguish what the system learned from where the behavior breaks down.

COOP-Transfer is independently developed and emphasizes prospective experimental design, reproducibility, explicit controls, frozen artifacts, and reporting negative results rather than optimizing the research narrative around positive findings.