Building Elucience. Researching how AI models form beliefs and how to intervene for better alignment. Sharing experiments, findings and prototypes.