Tomas Mikolov

Researcher best known for Word2Vec (Mikolov et al., 2013) — the CBOW and skip-gram word-embedding architectures with negative sampling that popularized dense word vectors and the king − man + woman ≈ queen embedding-algebra observation. Word2Vec’s CBOW scheme is reused directly by Word2Box, and its skip-gram/negative-sampling recipe underpins many learned search embeddings across this vault (Query2Vec, listing embeddings, synonym mining).

Key Contributions

• Word2Vec — CBOW & skip-gram

Articles

• Express Words in a Box - Understanding Box Embedding from the Basics — Word2Box builds on Mikolov’s CBOW