Multi-step reasoning in Prompt Engineering / GenAI - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding Multi-step Reasoning in AI

Which of the following best describes multi-step reasoning in AI?

A AI performing a sequence of logical steps to reach a conclusion

B AI randomly guessing answers without following any process

C AI only memorizing data without any reasoning

D AI making a single decision based on one input without intermediate steps

💡 Hint

Think about how humans solve problems step-by-step.

Multi-step reasoning means the AI breaks down a problem into smaller steps and solves each step logically to reach a final answer.

❓ Predict Output

intermediate

Output of Multi-step Reasoning Code

What is the output of the following Python code simulating multi-step reasoning?

Prompt Engineering / GenAI

def multi_step_reasoning(x):
    step1 = x + 2
    step2 = step1 * 3
    step3 = step2 - 4
    return step3

result = multi_step_reasoning(5)
print(result)

A 19

B 23

C 21

D 17

💡 Hint

Calculate each step carefully: add 2, multiply by 3, then subtract 4.

Step 1: 5 + 2 = 7
Step 2: 7 * 3 = 21
Step 3: 21 - 4 = 17
Output is 17.

❓ Model Choice

advanced

Choosing a Model for Multi-step Reasoning Tasks

You want to build an AI that solves math word problems requiring multiple logical steps. Which model type is best suited?

A A transformer-based language model with attention mechanisms

B A convolutional neural network (CNN) for image recognition

C A simple linear regression model

D A k-nearest neighbors (KNN) classifier

💡 Hint

Think about models good at understanding sequences and context.

Transformer models with attention can handle sequences and context well, making them suitable for multi-step reasoning in language tasks.

❓ Hyperparameter

advanced

Hyperparameter Impact on Multi-step Reasoning Models

Which hyperparameter adjustment is most likely to improve a model's ability to perform multi-step reasoning?

A Reducing the number of layers in a deep neural network

B Increasing the model's attention heads in a transformer architecture

C Using a higher learning rate without any decay

D Decreasing the batch size during training

💡 Hint

More attention heads help the model focus on different parts of the input simultaneously.

Increasing attention heads allows the model to capture more complex relationships, aiding multi-step reasoning.

🔧 Debug

expert

Debugging Multi-step Reasoning Model Output

A multi-step reasoning model outputs the same answer for all inputs. What is the most likely cause?

A The model is overfitting the training data

B The loss function is not decreasing during training

C The model's output layer weights are not updating due to a frozen layer

D The input data is too diverse and complex

💡 Hint

Think about why the model might produce constant outputs regardless of input.

If the output layer weights are frozen, the model cannot learn to differentiate inputs, causing constant outputs.