Embeddings — Learn free

Self-check

Questions 1–10 of 10

1. What are embeddings?
Question 1A. Numerical representations of data that capture meaning, context or relationships in a form machines can understandB. Compressed copies of raw text filesC. Encrypted versions of imagesD. Relational tables that store words as symbolsE. Random numbers assigned to each record
2. In the context of embeddings, what is a vector?
Question 2A. A type of database indexB. A list of numbersC. A keyword search queryD. A compressed image fileE. A rule for classifying text
3. Which pair of words is given as an example of embeddings that would be close to each other in vector space?
Question 3A. Apple and carB. Fish and birdC. King and queenD. Dog and fishE. Cheap and expensive
4. In data architecture, where are embeddings often stored so that similarity search can be performed?
Question 4A. In spreadsheetsB. In relational tables of raw textC. In email archivesD. In vector databasesE. In object storage as image files
5. How does an embedding-based search differ from a traditional keyword search when a user searches for 'cheap running shoes'?
Question 5A. It returns only exact matches of the phraseB. It ignores the query and shows popular productsC. It translates the query into another languageD. It searches only product imagesE. It understands that 'affordable sneakers' may mean the same thing even though the words differ
6. In the Grade 7 animal example of embeddings, why are dog and wolf considered similar?
Question 6A. Because their number sets, such as [0.8, 0.6, 0.7] and [0.82, 0.59, 0.72], are close to each otherB. Because they have the same name lengthC. Because they both live in waterD. Because their numbers are as different as possibleE. Because fish and birds were placed next to them
7. Conceptually, what does each dimension of an embedding's multi-dimensional space capture?
Question 7A. A single letter of the original wordB. A certain pattern learned from large amounts of dataC. The file size of the original dataD. The date the data was createdE. The name of the storage system
8. Which of the following is NOT listed as a use of embeddings?
Question 8A. Similarity searchB. ClusteringC. Encrypting passwordsD. ClassificationE. Retrieval-based generation
9. What does an image embedding capture, compared with an audio embedding?
Question 9A. Both capture the file nameB. An image embedding captures sound characteristics; an audio embedding captures visual patternsC. Neither captures anything beyond raw pixel or sample valuesD. An image embedding captures visual patterns; an audio embedding captures sound characteristicsE. Only text can have embeddings
10. How do embeddings enable AI systems to combine text and image understanding?
Question 10A. By storing both as plain keywordsB. By converting images into text descriptions onlyC. By keeping each modality in a separate system that never interactsD. By ranking images by file sizeE. By converting different modalities into vectors so systems can compare them mathematically