Text embedding models in Prompt Engineering / GenAI - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

What is the main purpose of text embedding models?

Text embedding models convert text into numbers. What is their main purpose?

A To transform text into fixed-length vectors that capture semantic meaning

B To translate text from one language to another

C To generate new text based on a prompt

D To count the number of words in a sentence

💡 Hint

Think about how computers understand text in a way that machines can work with.

Text embedding models create fixed-length vectors that represent the meaning of text, enabling machines to compare and analyze text effectively.

❓ Model Choice

intermediate

Which model is best suited for generating text embeddings?

You want to create embeddings for sentences to compare their meanings. Which model type is best?

A Transformer-based model trained on large text corpora

B Recurrent Neural Network (RNN) trained on time series data

C Convolutional Neural Network (CNN) trained on images

D Generative Adversarial Network (GAN) for image generation

💡 Hint

Consider models designed to understand language context deeply.

Transformer-based models like BERT or GPT are trained on large text datasets and excel at capturing semantic relationships in text, making them ideal for embeddings.

❓ Predict Output

advanced

What is the output shape of embeddings for 3 sentences using a model that outputs 768-dimensional vectors?

Given this Python code snippet:

sentences = ["Hello world", "Machine learning is fun", "AI helps humans"]
embeddings = model.encode(sentences)
print(embeddings.shape)

What will be printed?

Prompt Engineering / GenAI

sentences = ["Hello world", "Machine learning is fun", "AI helps humans"]
embeddings = model.encode(sentences)
print(embeddings.shape)

A (768, 3)

B (3, 768)

C (3,)

D (768,)

💡 Hint

Each sentence gets a vector of length 768. How many sentences are there?

The model encodes each sentence into a 768-dimensional vector. Since there are 3 sentences, the output shape is (3, 768).

❓ Metrics

advanced

Which metric is best to measure similarity between two text embeddings?

You have two text embeddings and want to measure how similar their meanings are. Which metric is most appropriate?

A Euclidean distance

B Accuracy

C Mean squared error

D Cosine similarity

💡 Hint

Think about a metric that measures the angle between two vectors rather than their length.

Cosine similarity measures the angle between two vectors, focusing on their direction, which is ideal for comparing semantic similarity of embeddings.

🔧 Debug

expert

Why does this embedding code raise a TypeError?

Consider this code snippet:

text = "AI is amazing"
embedding = model.encode(text)
print(embedding.shape)

It raises: TypeError: 'float' object has no attribute 'shape'. Why?

A The variable 'text' is not defined

B The model.encode returns a float instead of a vector

C The model.encode expects a list of strings, not a single string

D The print statement is missing parentheses

💡 Hint

Check the input type expected by the encode method.

Many embedding models expect a list of strings to encode multiple texts. Passing a single string returns a vector, but sometimes it returns a float if the model is misused. Here, the error suggests the input should be a list like [text].