Sampling is how a language model picks one of several possible next words. It decides how predictable or varied texts become.
Example
With narrow sampling, product texts sound uniform; with wide sampling you get more creative but less predictable ideas.
For pros
Top-k picks only from the k most likely tokens; top-p (nucleus sampling) from the smallest set whose combined probability reaches p. Together with temperature this controls variety. For reasoning models these settings are often fixed.
