DSA Studio

Embeddings Checkpoint

Dense representations and similarity.

Get every question correct to pass this checkpoint. You can review feedback, revise your answers, and submit again.

1. An embedding table has shape (8,3); IDs [2,2,5] are looked up. What are output shape and parameter count?

Select one answer.

2. Loss sums all lookup coordinates for IDs [4,4,1]. What gradient does row 4 receive from this lookup?

Select one answer.

3. A co-purchase model places printers near ink. What does that establish?

Select one answer.

4. What is cosine similarity between [1,0] and [2,0]?
5. A static embedding passes some king/man/woman analogies. Which conclusion is justified?

Select one answer.

Join Discord