Thread

Decision models like Jev don't beat LLM-as-a-judge or traditional classifiers

96 points · 34 comments · tomncooper

prodigycorp · 1791196285 · focus · HN ↗
The amount of run that Laya has gotten in this hype cycle when laya was nothing like jev is one heck of a phenomena. Do people even realize he took a finetuned bert model, gave it a jev-like head (after jev was released), and pushed the model to the public after jev was released, thus invalidating his claim of being copied?

Unofficial Hacker News client; not affiliated with Y Combinator.