Currently the solver does not use the previous values of inputs at all when solving. But it could use it in some cases as a heuristic I guess, yes!