Prefill

The first pass, where the whole prompt is processed at once to populate the KV cache. It determines time-to-first-token; the per-token decoding that follows is much cheaper.

Companion explanation in Step by Token, chapter 18.