Concept
Greedy decoding
Greedy decoding is the simplest strategy for turning a language model's output distribution into actual text: at each step, pick the single highest-probability token, append it, and repeat. Like a greedy graph traversal, it commits to the locally best choice without…
The rest of “Greedy decoding” is a premium feature: every concept in the library gets a precise, practitioner-focused write-up like this one, cross-linked straight from the paper summaries that use it.
Log in to unlock→