Can a Baby Learn? Beliefs and Certainty
The same father, chair, and child from Part I, now holding Beta beliefs: consistency, harsh punishment, and a new task, learned through credibility-weighted updates.
Blog
Notes on modelling, inference, and the mathematics behind the methods.
Can a Baby Learn? A Toy Model of Discipline and Desire
A father, a chair, and a Q-learning agent: what a toy reinforcement-learning model says about consistency, harsh punishment, and a child’s desire to climb.