'system has learned from the submission / mail is automatically allowed'

Hey folks, got an alert about a tenant allow//block list entry expiring. Only recently did we start getting these, because only recently did we start using expiring whitelisting. But I'm a little confused by the details, which says 'Mail from x is now automatically alllowed and the allow entry has been removed' and the activity that ''an allow entry is no longer required as the system has learned from the submission'

The referenced email is actually an internal tenant - it receives ticket requests, and sends out ticket updates. But I'm REALLY curious about the 'automatic' allowing. Is this a feature limited to Defender 2, or part of Microsoft's AI detection framework for all 365 Defender/EOP? I don't even remember submitting this email - if I did, it was probably more than 45 days ago. So 

1) Is this notice primarily that the entry had expired, but ALSO it's not needed or does this send out as soon as 'the system' recognizes it  as legitimate, and removed regardless of the time left? 

2) is there a way to review a list of entries Microsoft has 'accepted'?

3) What exactly does this 'allow'? I know that the tenant allow/block list allowed a certain set of lower-risk indicators in an email, but still blocked some higher-risk ones - unless there was a submission made. At that point, more is allowed. But there's still a limit, compared to a blanket bypass on the policy itself.