@kiki_morozova's Black Hat Europe talk on our image scaling research is on YouTube now! If you liked Anamorpher, check it out for more cool image preprocessing exploits-- plus some fun audio ones too!
I started playing with Jev. I wanted to see how hard it is to find adversarial examples for it. For about a dollar, I found a poem that the model classifies as "code" rather than "language".
There are glaring issues in this paper beyond the evaluation problems raised by Davi and ToB, including ones that make me question the ratio of human to AI assistance here, but a few are so egregious we should consider what research norms we are demanding from industry labs.