Review labels and comments alike are copied to the analysis results of a subsequent scan.
Once a review label was added to a vulnerability, the same vulnerability in upfollowing scans automatically inherits this review label to save rework.
By using matching algorithms, this works even if the code of the vulnerability or the one surrounding it changes to some extend.
"Not exploitable", "No issues", and "Duplicate" are negative reviews, i.e. the issue is not exploitable for one reason or another.
Issues with negative reviews are ignored by the RIPS CI tools. They are also hidden from the UI by default (this can be changed in the account settings).