i only briefly started playing with this (so nice to be able to run this locally on my framework lol), but the one thing that i am impressed with is that this is a thinking model, so getting to see the reasoning for why the model believes content is/isnβt violative is great
RE:
things have definitely gotten a bit worse, but recently have been putting a considerable amount of work into cleaning up some of the recurring spam problems. suspect that you'll start seeing improvements over the next few days. please do report things. even though it might seem useless, it does help