Synthetic Data

Why does using the same model family as synthetic-data teacher AND eval judge inflate FT scores?

When eval prompts appear verbatim inside an SFT file after a synthetic-data run, what's the root cause?

Which operators are part of the Evol-Instruct pipeline for growing instruction data?

Where does quality leak in a self-instruct synthetic data pipeline?

Distillation with a closed-API teacher: which method is feasible?