A static, word2vec-style embedding is trained on a large corpus. The vocabulary was fixed before training began and holds a single entry for the token bank, which occurs times:
| context | occurrences | words typically nearby |
|---|---|---|
| finance | money, cash, account, loan | |
| river | river, water, mud, fishing |
The finance words and the river words occupy two well-separated regions of the embedding space. Training runs to convergence.
Which statement best describes the bank row of the table once training has finished?
Select all that apply.