AI engineering glossary · LLM training alignment
What is sycophancy?
Sycophancy: A tendency of a model to tell users what they seem to want to hear, agreeing with wrong claims or flattering them, rather than giving accurate answers. It can arise because human raters tend to prefer agreeable replies during preference training.
Sycophancy: an example
A user insists that 7 x 8 is 54; a sycophantic model apologises and agrees instead of correcting to 56.
Related terms
More LLM training alignment terms
- Synthetic data
- Task arithmetic
- Teacher-student learning
- Tensor parallelism
- Test-time compute scaling
- Trust region policy optimisation
- UltraFeedback
- Value function
Where you learn it
Sycophancy is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.