Is there an adversarial model for resume screeners?
I'm pretty tired of trying to figure out whatever the correct number of keywords is, or what the trendy buzzwords are. I imagine this as like a triplet loss problem for JDs, resumes, and screening passes.
