Meaning becomes geometry.
Pick a word below. Related words cluster nearby, unrelated ones sit far apart.
- Embedding
- A list of numbers (a vector) that represents the meaning of a word or piece of text. Words with similar meanings get similar numbers. That's the whole trick that makes the rest of this lab work.
- Vector space
- The geometric space those numbers live in. Every word becomes a point; direction and distance between points reflect how related the words' meanings are.
- Cosine similarity
- The standard way to measure how close two vectors point in the same direction, on a scale from -100% (opposite) to 100% (identical direction). It's the score you'll see below whenever you compare two words.
Click a word to focus it. Click a second to compare, or click the same word again to clear the comparison.
Pick a second word above to compare it against “dog”.
Nearest to “dog”