💬Reward Model
Also known as Preference Model, RM
A model that scores answers the way people would. It guides training.
Think of it like
Think of it like a judge at a cake show. They give each cake a score.
Example
Emma's team trains a reward model on 80,000 picks. It now gives a clear, kind reply 0.9 and a rude one 0.2.
Why it matters
It lets a few thousand human choices guide millions of training steps.
Definition
Common mistake
Key takeaways
- 1Scores answers. Like a human judge.
- 2Learns from picks. Better of two, many times.
- 3Can be gamed. Blind spots get exploited.