Marginalized posterior distribution
like getting a good idea of the average weight
Technical definition
A probability distribution over the model parameters, marginalized over the other variables, which gives the probability distribution of a single parameter.