Just to compare what was posted at LLO yesterday https://alog.ligo-la.caltech.edu/aLOG/index.php?callRep=12702.
I ran 128k 3Hz sinewave to each quadrant with -125k bias on the DC and looked at the responses.  The good news is it actually responds.  The bad new it trying to make it respond to sensible signals like a pitch or a yaw does not yield much in a way a results.  I am sure Keita will have more to say.  But attached are our results of a 3Hz injection