TRANSFORMERS & LLMs • 73 / 397
Follow text from characters to tokens, attention, transformers and generation.

Top-p

Top-p limits token selection to a probability mass containing the most likely candidates before sampling.

Think of it like

Think of Top-p as part of a very fast reader that repeatedly decides what information matters and what should come next.

Real life

Autocomplete, translation and chat assistants depend on concepts such as Top-p.

SRE lens

It is another decoding control alongside temperature.

Remember thisTop-p limits token selection to a probability mass containing the most likely candidates before sampling.
AIForSREJump to a concept
Search is optional. The main journey is simply ↓ Next.