**Part 1: Extrapolating from smooth functions.**

In the first experiment, described in Section 4.2 of the paper, participants were asked to exptrapolate from several functions, where the true underlying relationships were draws from a Gaussian process with a rational quadratic kernel. [Link]**Part 2: Extrapolating from smooth functions.**

In the second experiment, described in Section 4.3 of the paper, participants were asked to exptrapolate from functions that are difficult or impossible for conventional Gaussian process methods to capture. [Link]**Part 3: Preference for smoothness/simplicity.**

In the third experiment, described in Section 4.4 of the paper, participants were asked to express their preferences over different kinds of explanations or underlying relationships, given a small number of data points. [Link]