On further thought, it looks like I've been garbling my nomenclature; so for the record:
yaw rotation about the normal/vertical axis,
sideslip angle airflow makes with the longitudinal axis.
So the yaw damper and/or rudder program(s) yaw the a/c to zero out sideslip.
And sideslip pretty much has to be sensed by an airflow sensor, be it a vane or a differential pressure between static ports or other opposite side ports in the fuselage or fin (A sensor using a classic ball would surprise me, but I'm not an instrumentation engineer).
As different flow regions are entered and exited within a vortex, the sideslip sensor could be presenting highly variable and possibly misleading sideslip conditions to the rudder program which can result in the yaw damper amplifying the condition rather than dampening it. And possibly it could get caught in FCSIO, the computer equivalent of PIO.
Now if your a/c suddenly yaws heavily, you may likely put in full opposite rudder, but don't forget the NTSB has just published that so doing may break it off <img src="frown.gif" border="0">
You may prefer to turn off the yaw damper and any other automatic rudder control immediately upon entering a wake vortex.