AI0-001 Practice Test 2 — 10 Questions

Question 1

An AI team is deploying a fine-tuned LLM for a code generation assistant. They need to ensure the model outputs only syntactically valid JSON for integration with downstream systems. Which prompt engineering technique is MOST effective for enforcing structured output?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Enable JSON mode in the API call, specifying the desired JSON schema. Enabling JSON mode in the API call and specifying the desired JSON schema is the most effective technique because it constrains the model's decoding process at the API level, forcing the output to conform to valid JSON structure. This is a hard constraint enforced by the inference engine, not a soft suggestion in the prompt. It eliminates the risk of malformed output that downstream systems cannot parse.

Answer

Provide a few-shot example of a valid JSON response in the prompt

Answer

Include a system prompt that says 'You are a helpful coding assistant.'

Answer

Use chain-of-thought prompting to have the model reason step-by-step before answering

Question 2

A fraud detection model has high precision but low recall. The cost of false negatives is very high. Which threshold adjustment should be made?

Accepted Answer

Decrease classification threshold. Decreasing the classification threshold makes the model more sensitive, classifying more instances as positive. This increases recall by catching more true positives, directly addressing the high cost of false negatives, even though precision may drop.

Answer

Use class weights during training

Answer

Apply SMOTE to the training data

Answer

Increase classification threshold

Question 3

An AI developer is building an agent that can book flights and hotels by calling external APIs. The agent needs to decide which API to call and in what order based on user requests. Which pattern is BEST suited for this multi-step reasoning and tool use?

Accepted Answer

Apply the ReAct pattern (Reasoning and Acting). The ReAct (Reasoning and Acting) pattern interleaves natural-language reasoning traces with tool/API calls, allowing the agent to decide which API to invoke, observe the result, and reason about the next step. This iterative loop is ideal for multi-step tasks like booking flights and hotels, where the sequence depends on intermediate results (e.g., flight availability before hotel dates). It is the best-suited pattern for dynamic, multi-step tool use.

Answer

Implement a simple Retrieval-Augmented Generation (RAG) pipeline

Answer

Fine-tune a model to output API call sequences directly

Answer

Use function calling with a fixed sequence of API calls

Question 4

In the AI project lifecycle, after a model is trained and evaluated, it is deployed to a production environment. What is the NEXT critical step to ensure the model continues to perform well over time?

Accepted Answer

Monitoring the model's performance and data drift. After deployment, the next critical step is monitoring the model's performance and detecting data drift, because production data distributions change over time and model accuracy degrades silently. Monitoring closes the MLOps loop by feeding real-world performance signals back to the team, enabling timely retraining or rollback. Without monitoring, degradation goes unnoticed until business impact occurs.

Answer

Collect more training data

Answer

Archive the model and start a new project

Answer

Re-train the model from scratch

Question 5

A company uses an AI model to screen job applicants. A disparate impact analysis reveals that the model's rejection rate for a protected group is significantly higher than for others. Which THREE actions should the company take to address this?

Accepted Answer

Revisit training data for historical bias and consider reweighting. Revisiting the training data for historical bias and applying reweighting directly addresses the root cause of disparate impact. If the training data contains biased labels or skewed representation of the protected group, the model will learn and amplify those biases. Reweighting adjusts the loss function to give more importance to underrepresented or disadvantaged groups, helping to equalize error rates across groups.

Answer

Ignore the disparity because the model is accurate overall

Answer

Remove all demographic attributes from the dataset

Question 6

An AI system that can perform any intellectual task that a human being can is referred to as:

Accepted Answer

Artificial General Intelligence (AGI). Artificial General Intelligence (AGI) refers to a hypothetical AI system that possesses the ability to understand, learn, and apply intelligence across a wide range of tasks at a level comparable to a human being. It contrasts with narrow AI, which is specialized for specific tasks. The definition in the question matches AGI exactly.

Answer

Machine learning

Answer

Narrow AI

Answer

Deep learning

Question 7

An engineer is building a regression model to predict housing prices. The dataset includes features such as square footage, number of bedrooms, and year built. The engineer notices that the square footage values range from 500 to 10,000, while the number of bedrooms ranges from 1 to 5. Which preprocessing step is most critical before training a gradient descent-based model?

Accepted Answer

Normalize or standardize the features. Gradient descent-based models are sensitive to the scale of input features because they update weights proportionally to the gradient, which is influenced by feature magnitudes. With square footage ranging 500–10,000 and bedrooms 1–5, the larger feature will dominate the gradient, causing slow or unstable convergence. Normalizing or standardizing (e.g., Z-score or min-max scaling) ensures all features contribute equally, leading to faster and more reliable training.

Answer

Use k-fold cross-validation

Answer

Apply log transformation to all features

Answer

One-hot encode the features

Question 8

A machine learning engineer is tuning a neural network for image classification. The training loss decreases steadily, but the validation loss starts increasing after 50 epochs. Which action best addresses this issue?

Accepted Answer

Apply early stopping with a patience of 10 epochs. The described behavior—decreasing training loss with increasing validation loss—is a classic sign of overfitting. Early stopping with a patience of 10 epochs directly addresses this by halting training when the validation loss fails to improve for a specified number of epochs, preventing further overfitting while retaining the best model weights.

Answer

Increase the number of hidden layers

Answer

Add more training data

Answer

Increase the batch size

Question 9

An AI practitioner is fine-tuning a large language model for a domain-specific task using a small labeled dataset (500 examples). They have limited GPU memory. Which technique is MOST suitable?

Accepted Answer

QLoRA (Quantized Low-Rank Adaptation). QLoRA (Quantized Low-Rank Adaptation) is the most suitable technique because it combines 4-bit quantization of the base model with low-rank adapter modules, drastically reducing GPU memory usage while still allowing fine-tuning on a small dataset. This approach preserves the model's pre-trained knowledge and avoids catastrophic forgetting, which is critical when only 500 labeled examples are available.

Answer

Full fine-tuning of all model parameters

Answer

Instruction tuning with the full dataset

Answer

Retrieval-Augmented Generation (RAG) without fine-tuning

Question 10

Which THREE are effective methods for ensuring data privacy in AI training? (Choose three.)

Accepted Answer

Data anonymization. Data anonymization (B) is correct because removing or masking personally identifiable information (PII) such as names, addresses, and identifiers before training prevents the model from learning or exposing individual identities. Differential privacy (C) is correct because it adds calibrated statistical noise (e.g., via mechanisms like Laplace or Gaussian noise with a privacy budget epsilon) so that the inclusion or exclusion of any single record has a negligible effect on outputs, providing a formal privacy guarantee. Federated learning (E) is correct because it trains models locally on each device or silo and shares only model updates (e.g., gradients or weights) rather than raw data, keeping sensitive data on the originating endpoint. Data encryption at rest (A) protects stored data against unauthorized access but does not prevent privacy leakage during training or inference, so it is not one of the three methods for ensuring privacy in AI training. Data replication (D) merely copies data to additional locations, which increases exposure and does nothing to protect privacy, so it is not a valid method.

Answer

Data encryption at rest

Answer

Data replication