Pinned
When do language models moderate content well, and where do they struggle? Bring a set of rules you know well, and we’ll test them across open models in our session with Musubi! All T&S folks welcome, no ML background needed! Next Tuesday, Oct 27. Sign up here: