Recently, I was notified of this error in the alert log of one of my Oracle 18.104.22.168 databases:
ORA-00600: internal error code, arguments: , , , [memory leak], , , , , , , , 
With the words “memory leak”, as part of the error, I reached out to Oracle support.
After Oracle Support evaluated the trace file I sent associated with this error, they replied: “can be safely ignored – it is just an internal housekeeping error and the size involved (xx bytes) is small”
The Oracle Support recommend:
“You can prevent the error from being reported by setting event 10262 in the init.ora or spfile and restarting the instance. eg: Add this parameter setting to either the init.ora or spfile: event = “10262 trace name context forever, level 2000000”
alter system set event= ‘10262 trace name context forever, level 2000000’ scope=spfile;
In either case restart the instance so that all processes use the new setting.”