José Wheeler

Independent AI Safety researcher and two-time medalist in the Argentine Chemistry Olympiad. My work focuses on Mechanistic Interpretability, specifically using Sparse Autoencoders to audit internal reasoning circuits in LLMs. Beyond the technical side, I am a high-agency student preparing to pursue a Cognitive Science Major in the US (2027), driven by a deep fascination with how mental frameworks apply to artificial systems. I am dedicated to bridging the gap between Latin American talent and global AI Safety efforts, transforming technical curiosity into rigorous research that ensures a transparent and safe future for AI.

Personal Website

-

LinkedIn

-

Twitter / X

-

LessWrong

-

EA Forum

-

Projects

No grants recorded.