Decision models like Jev don't beat LLM-as-a-judge or traditional classifiers (developers.redhat.com)

<a href="https://news.ycombinator.com/item?id=49933476">Comments</a>