Word2Vec
Last Updated: July 29, 2026 | By Mihail Sebastian | AI Dictionary
A 2013 Google method that learns word embeddings by training a neural network to predict a word from its context – famous for king − man + woman ≈ queen.
What is Word2Vec?
Word2Vec is an algorithm that learns word embeddings by training a small neural network to predict a word from the words around it. It was published by Tomas Mikolov and colleagues at Google in 2013.
Word2Vec is one specific method; the word embedding is the general concept it produces. Its results made the idea famous: meaning, learned purely from word co-occurrence, behaves like geometry.
How Word2Vec Works
Word2Vec comes in two variants. Continuous Bag of Words (CBOW) predicts a word from its surrounding context; Skip-gram does the reverse and predicts the surrounding words from a single target word, which tends to give better vectors for rare words.
The prediction task is scaffolding. After training, the predictor is thrown away and the network’s learned weights are kept – those weights are the embeddings. The network itself is shallow, a single hidden layer, which is why it trains quickly even on very large corpora.
The logic is simple: to predict neighbors well, the network must assign similar vectors to words used in similar contexts. “Cat” and “dog” show up around the same words, so they end up close together.
Example of Word2Vec
Train Skip-gram on a large news corpus. A window slides over “the cat sat on the mat” and generates training pairs: from the target “sat”, the network must predict “cat”, “on”, and the rest of the window.
Each pair nudges the vectors slightly. Repeat across billions of words, and structure accumulates: animals cluster together, verb tenses line up, capital cities sit in a consistent relation to their countries.
Mikolov’s team reported the result that made the method famous: taking the vector for “king”, subtracting “man”, and adding “woman” lands nearest to “queen”. Analogies had become arithmetic.
Related AI terms: Word Embedding · Embedding Layer · Tokenization · Natural Language Processing
Did you like the Word2Vec gist?
Learn about 250+ need-to-know artificial intelligence terms in the AI Dictionary.
Mihail Sebastian — Writes about AI governance, regulation, and the technology behind them. Placeholder bio — replace with a real credential line. About