Hacker News new | ask | show | jobs
by whattheheckheck 15 days ago
Thats when your build a better Ralph loop around your llm for it to converge to an answer and not rely on 1 shots