‹ BackHN Continuity

Thread

LLM Classification Is Feature Engineering

116 points · 25 comments · minsufficient

  1. softwaredoug · · focus · HN ↗
    In my work on LLM as a judge, I prefer to use LLM decisions as features in a downstream classic ML model for the final decision. It works really well

    <a href="https:&#x2F;&#x2F;softwaredoug.com&#x2F;blog&#x2F;2025&#x2F;01&#x2F;21&#x2F;llm-judge-decision-tree" rel="nofollow">https:&#x2F;&#x2F;softwaredoug.com&#x2F;blog&#x2F;2025&#x2F;01&#x2F;21&#x2F;llm-judge-decision-...

    1. chrisweekly · · focus · HN ↗
      Cool approach, thanks for sharing.

      FWIW I also liked this (very different) post of yours: <a href="https:&#x2F;&#x2F;softwaredoug.com&#x2F;blog&#x2F;2026&#x2F;07&#x2F;13&#x2F;who-want-to-be" rel="nofollow">https:&#x2F;&#x2F;softwaredoug.com&#x2F;blog&#x2F;2026&#x2F;07&#x2F;13&#x2F;who-want-to-be

      1. softwaredoug · · focus · HN ↗
        Thank you!
Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.