Content writing assistance in Prompt Engineering / GenAI - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

How does a language model generate text?

Imagine a language model trained to write stories. How does it decide the next word to write?

A It predicts the next word based on the previous words it has generated.

B It randomly picks any word from its vocabulary without context.

C It always repeats the first word it learned during training.

D It selects the longest word available in the vocabulary.

💡 Hint

Think about how sentences flow naturally when you speak or write.

Language models generate text by predicting the next word based on the words that came before, using patterns learned from training data.

❓ Predict Output

intermediate

Output of simple text generation code

What is the output of this Python code simulating a simple text generation step?

Prompt Engineering / GenAI

import random
words = ['hello', 'world', 'machine', 'learning']
context = ['machine']
next_word = random.choice(words)
print(next_word)

A Raises a TypeError

B Prints 'hello world' concatenated

C Always prints 'machine'

D Prints a random word from the list including 'machine'

💡 Hint

Look at how random.choice works on a list.

random.choice picks one element randomly from the list, so output can be any word from the list.

❓ Model Choice

advanced

Choosing a model for content writing assistance

You want to build a content writing assistant that can generate long, coherent articles. Which model type is best?

A A large transformer-based language model trained on diverse text data

B A clustering model grouping similar sentences

C A simple linear regression model predicting word counts

D A small decision tree model trained on article titles

💡 Hint

Think about models that understand language context deeply.

Transformer-based language models are designed to generate coherent and context-aware text, suitable for long articles.

❓ Metrics

advanced

Evaluating content writing model quality

Which metric is most appropriate to evaluate how well a content writing model generates fluent and relevant text?

A Accuracy of classifying text sentiment

B Mean Squared Error (MSE)

C BLEU score comparing generated text to reference text

D Number of words generated per second

💡 Hint

Think about metrics that compare generated text to human-written text.

BLEU score measures how similar generated text is to reference text, useful for evaluating text generation quality.

🔧 Debug

expert

Debugging unexpected repetitive output in text generation

A content writing model keeps repeating the same phrase over and over. What is the most likely cause?

A The optimizer learning rate is too high causing slow learning.

B The model's temperature parameter is set too low, causing low randomness.

C The model has too many layers causing overfitting.

D The training data was shuffled randomly before training.

💡 Hint

Consider how randomness affects text diversity.

A low temperature makes the model pick the highest probability word repeatedly, causing repetitive output.