Comments

Nice work on this. One thing I'm curious about: what's the refresh cycle for the prompt set behind the visibility metric? Given how much model answers can vary across runs, how do you handle that noise when tracking changes over time?