Instruction formatting in Prompt Engineering / GenAI - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding Instruction Formatting Impact

Which of the following best describes why clear instruction formatting is important when using generative AI models?

A It guarantees the model will always produce error-free code regardless of input.

B It reduces the model's training time significantly by skipping data preprocessing.

C It helps the model understand the task better, leading to more accurate and relevant outputs.

D It allows the model to generate outputs without any need for user input.

💡 Hint

Think about how clear instructions affect the model's understanding.

Clear instruction formatting guides the model to focus on the right task, improving output quality. It does not affect training time or guarantee error-free outputs.

❓ Predict Output

intermediate

Output of Instruction with Ambiguous Formatting

What will be the output of the following prompt given to a generative AI model? "Generate a Python function that adds two numbers and returns the result. Use clear variable names." Assuming the model follows instructions literally, what is the likely output?

Prompt Engineering / GenAI

def add_numbers(a, b):
    return a + b

A

def add_numbers(a, b):
    return a + b

B

def add(x, y):
    return x + y

C

def addNumbers(a,b):
    return a+b

D

def add_two_numbers(num1, num2):
    return num1 + num2

💡 Hint

Look for the exact variable names used in the prompt.

The prompt asks for clear variable names but does not specify them. The model likely uses generic names 'a' and 'b' as in option A, matching the prompt literally.

❓ Model Choice

advanced

Choosing Model Type Based on Instruction Complexity

You want to generate detailed step-by-step explanations for complex math problems using a generative AI. Which model type is best suited for this task?

A A model trained exclusively on code snippets without natural language understanding.

B A model specialized only in image recognition tasks.

C A small, fast model optimized for short text completions.

D A large language model trained on diverse datasets with strong reasoning capabilities.

💡 Hint

Consider the need for reasoning and detailed explanations.

Large language models trained on diverse data can understand complex instructions and generate detailed explanations, unlike small or specialized models.

❓ Hyperparameter

advanced

Effect of Temperature on Instruction Output

When generating text from a generative AI model, what is the effect of increasing the 'temperature' hyperparameter?

A The output becomes shorter and more concise.

B The output becomes more random and creative, with less predictable word choices.

C The model runs faster but with lower accuracy.

D The output is guaranteed to be grammatically correct.

💡 Hint

Think about randomness and creativity in text generation.

Higher temperature increases randomness, making outputs more creative but less predictable. It does not affect length, speed, or grammar guarantees.

❓ Metrics

expert

Evaluating Instruction Following with BLEU Score

You want to measure how well a generative AI model follows instructions by comparing its output to a reference answer. Which metric is most appropriate for this task?

A BLEU score, which measures overlap of n-grams between generated and reference text.

B Mean Squared Error, which measures numeric differences between outputs.

C Accuracy, counting exact matches only.

D Confusion matrix, showing true vs predicted classes.

💡 Hint

Consider metrics used for comparing text similarity.

BLEU score is widely used to evaluate text generation quality by comparing n-gram overlaps, suitable for instruction following tasks.