The tokenization technique that uses white space while tokenizing is called the "Whitespace Tokenizer." This method splits text into tokens based on white space characters such as spaces, tabs, and new lines, treating each word separated by white space as a token without additional analysis like punctuation removal or lowercasing.
Other techniques mentioned, like word2vec, character-level tokenization, subword tokenization, and SentencePiece, do not primarily rely on whitespace but use different approaches such as splitting at character level, subwords, or using specific algorithms that consider spaces differently or ignore them as explicit delimiters in some cases.
Thus, whitespace is explicitly used in whitespace tokenization for segmenting text tokens.
