RAG embeddings vs Word2Vect

so when query from vector database, how a query being processed and linked to the chunks persisted in the database? so here the embedding is not exactly same as the term when used in word2vect context; word2vec = good at capturing word-level semanti...