There appears to be a temporary change in the persistence behavior of the LUCI detectors.  After an observation of Saturn, from data collected on 20220906 (simple sum of the dithered Saturn data), persistence is still visible a month later (20221013):

This effect is more prominent on the LUCI2 detector. LUCI1 is on the Left image above and LUCI2 on the right.  The feature manifests as a depressed persistence signal.

For LUCI2: The counts in the Saturn feature (~27ADU) are lower than just outside of it (~52 ADU).  This is much lower than the counts in the quarter-circle feature at bottom right (~100ADU) of the above image.  This is measured on a dark frame taken immediately after K band exposures averaging ~33k ADU (~60% full well).

There are some other structures on the dark exposure that look like trails from a bright star as the telescope moved around (e.g. curve to the left of the Saturn dice-5).  This feature should come out with normal data reductions.  It is suggested that people take calibrations in a timely fashion to science data taken.