# How much is one bit of information gain?

From Wikipedia on Bayesian experimental design, we take the definition of the utility function from Bayesian inference:
$U\left(\xi \right)=\int p\left(D|\xi \right)U\left(D,\xi \right)dy$