Training-method researcher
Taku Kudo
Created SentencePiece and unigram tokenization, the language-agnostic subword tokenizers used across modern models.
Works in the archive (2) · lead author on 2
- LeadSentencePiece · 2018
- LeadSubword Regularization / Unigram LM Tokenization · 2018
“Lead” marks the first-listed author — a byline fact, not a measure of influence; co-authorship is shown as recorded.