*todo: validate if improved presences are suposed to activate after learning or on next presence switching