I don't think it's based on your contributions. My main account had over 600k comment karma when it was banned.
They are purging anti-corporate perspectives and anti-genocide perspectives, this is becoming clear. They now have a social credit system where they track accounts that upvote posts which are later removed.
They probably want to make the AI training data a bit more pro corporate and right wing.
I'm pretty sure its their oligarch bro overlords attempting to purge the AI training dataset of any opinions they don't like.