Vector similarity metrics in Prompt Engineering / GenAI - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is the purpose of vector similarity metrics in machine learning?

Vector similarity metrics measure how alike two vectors are. They help compare data points, like checking if two images or texts are similar.

beginner

Explain Cosine Similarity in simple terms.

Cosine Similarity measures the angle between two vectors. If the angle is small, the vectors are similar. It ignores their length and focuses on direction.

beginner

What is Euclidean Distance and how does it relate to similarity?

Euclidean Distance is the straight-line distance between two points (vectors). Smaller distance means higher similarity, bigger distance means less similar.

intermediate

How does Jaccard Similarity work for comparing vectors?

Jaccard Similarity compares two sets by dividing the size of their overlap by the size of their union. For vectors, it measures how many features they share compared to total features.

intermediate

Why might you choose Cosine Similarity over Euclidean Distance?

Cosine Similarity focuses on direction, ignoring length, which is useful when magnitude varies but pattern matters. Euclidean Distance considers magnitude, which can be misleading if scale differs.

Which metric measures the angle between two vectors?

A Cosine Similarity

B Euclidean Distance

C Jaccard Similarity

D Manhattan Distance

Cosine Similarity calculates the cosine of the angle between two vectors to measure similarity.

If two vectors have a Euclidean distance of zero, what does that mean?

A They are identical

B They are completely different

C They have no features

D They have opposite directions

A Euclidean distance of zero means the vectors are exactly the same point in space.

Jaccard Similarity is best used for comparing:

A Angles between vectors

B Continuous numeric vectors

C Distances in 3D space

D Sets or binary vectors

Jaccard Similarity compares overlap between sets or binary vectors.

Which similarity metric ignores the length of vectors and focuses on direction?

A Euclidean Distance

B Hamming Distance

C Cosine Similarity

D Jaccard Similarity

Cosine Similarity measures the angle between vectors, ignoring their length.

A higher Euclidean distance between two vectors means:

A They are more similar

B They are less similar

C They have the same direction

D They have identical features

Greater Euclidean distance means vectors are farther apart and less similar.

Describe three common vector similarity metrics and when you might use each.

Think about what each metric focuses on: angle, distance, or overlap.

Explain why choosing the right similarity metric matters in machine learning tasks.

Consider how data features and scale affect similarity.