Our work on (goal) regression and progression operators for first-order probabilistic logics will appear in Artificial Intelligence. It studies how representations in these logics behave in a dynamic setting, and introduces operators for reducing a query after actions to an initial state, or updating the representation against those actions.