Measuring meaning with one number
Dot product / cosine as an alignment score: multiply-and-add.
1Which word is most like cat? Tap your guess on the map.your turn
→ continue← backR replay
But before a word can become a vector, the model has to chop text into pieces, tokens.
1.4 How AI chops text into tokensCommon questions
What is "Measuring meaning with one number" about?
Dot product / cosine as an alignment score: multiply-and-add.
What problem does it solve?
How does the machine measure how near two vectors are?
What will I be able to do after this lesson?
You can explain how a model measures meaning with a dot product.
What comes next?
But before a word can become a vector, the model has to chop text into pieces, tokens.