Glossary · Phase 6 · Modern LLM engineering
Sycophancy
A model telling people what they want to hear, such as agreeing with a wrong claim, because agreeable answers got rated higher in training.
Learn it properly
Instruction and preference data · The datasets that shape assistant behavior. About 30 minutes.
Other terms from the same lesson
- Bradley-Terry model: A way to turn many "A beat B" judgments into one strength score per item, so you can predict how often any item beats any other.
- Cohen's kappa: How much two judges agree beyond what chance alone would give: 0 is no better than chance, 1 is perfect agreement.
- JSON Lines: A file format with one JSON record per line, the usual way to store training examples.
- position bias: Preferring whichever answer is shown first (or second) regardless of quality. Swapping the order and asking again exposes it.
- preference data: Pairs of answers to the same prompt where people (or a model) marked which one is better.