AI0-001 Practice Exam 55 — 20 Questions

Question 1

A team is implementing an ML pipeline using a feature store. Which benefit does a feature store primarily provide in an AI operations context?

Accepted Answer

Consistency of feature computation between training and inference. A feature store ensures that feature engineering logic is stored, versioned, and reused consistently across both training and inference pipelines. This eliminates training-serving skew, a common cause of model degradation in production, by guaranteeing that the same transformations are applied to data regardless of when or where it is computed.

Answer

Automated scaling of inference endpoints

Answer

Real-time monitoring of model performance

Answer

Automatic model versioning and rollback

Question 2

A deployed NLP sentiment analysis model experiences a sharp decline in accuracy on customer reviews. The team has verified the input data format and pipeline are correct. Which THREE actions should be taken to diagnose and remediate? (Choose 3.)

Accepted Answer

Analyze recent user input for distribution shifts compared to training data.. Option A is correct because when a deployed NLP model's accuracy drops while the input format and pipeline are verified as correct, the most likely cause is data drift — the statistical distribution of recent user input has shifted away from the training distribution, so comparing recent inputs against training data (e.g., via feature/token distributions, embeddings, or KS tests) is the proper first diagnostic step. Option D is correct because reverting to a previously well-performing model version is a valid remediation and diagnostic tactic: it restores service quality immediately and, if the older version still performs well on the new data, it confirms the regression was introduced by the newer model rather than by the data itself. Option E is correct because concept drift — a change in the relationship between inputs and the target label (e.g., sentiment words taking on new meaning) — is a leading cause of accuracy decay in production NLP models, so a structured root cause analysis targeting concept drift (and distinguishing it from data drift) is essential to select the right fix. Option B is not appropriate as stated because blindly retraining with all available data, including potentially mislabeled or drifted recent data, can propagate the problem and does not diagnose the cause. Option C is not appropriate because adding synthetic data addresses data scarcity, not the verified accuracy decline, and synthetic data can introduce its own distributional biases without first identifying the root cause.

Answer

Immediately retrain the model with all available data.

Answer

Increase the size of the training dataset by adding synthetic data.

Question 3

A company is deploying a chatbot using a large language model. They want to mitigate the risk of prompt injection attacks. Which TWO measures should be implemented?

Accepted Answer

Implement input validation and sanitisation. Option A is correct because input validation and sanitisation directly strip or neutralise adversarial payloads (e.g., embedded instructions, delimiter-breaking characters, or encoded jailbreak strings) before they reach the LLM, reducing the attack surface for prompt injection. Option B is correct because a strict system prompt establishes immutable behavioural boundaries and instruction hierarchy, making it harder for user-supplied text to override the model's intended role or exfiltrate system instructions. Option C is not a reliable mitigation because fine-tuning on safe examples does not prevent novel injection payloads at inference time and can even be undone by adversarial prompting. Option D is irrelevant, since a larger context window merely allows more tokens to be processed and does not filter or constrain malicious instructions. Option E only caps response size and does nothing to stop an injected prompt from altering the model's behaviour or leaking data.

Answer

Fine-tune the model on safe conversational examples

Answer

Use a larger context window

Answer

Limit the maximum output token length

Question 4

An organization wants to automate the detection of defective products on an assembly line using computer vision. They have a limited number of labeled images for defective items. Which approach would be most effective?

Accepted Answer

Use transfer learning with a pre-trained model like ResNet and fine-tune on the defect data. Transfer learning with a pre-trained model like ResNet is most effective because it leverages features learned from large datasets (e.g., ImageNet) and adapts them to the defect detection task with limited labeled data. Fine-tuning only the later layers preserves general visual features while specializing for defect classification, avoiding overfitting that would occur with a small dataset.

Answer

Use a support vector machine with handcrafted features

Answer

Train a convolutional neural network from scratch on the limited data

Answer

Synthesize additional defective images using GANs

Question 5

A bank uses an AI model to approve loans. During an audit, it is found that the model denies loans at a higher rate for a certain ethnic group. Which governance principle is primarily violated?

Accepted Answer

Fairness. The model's disparate impact on a specific ethnic group directly violates the principle of Fairness, which requires that AI systems do not discriminate based on protected attributes such as race, ethnicity, or gender. In lending, fairness is often assessed using metrics like demographic parity or equal opportunity, and a higher denial rate for one group indicates a lack of algorithmic fairness.

Answer

Accountability

Answer

Transparency

Answer

Privacy

Question 6

A data engineer is preparing a dataset for a binary classification model. The dataset has 10,000 samples with 100 features. To improve model performance and reduce training time, the engineer decides to perform feature selection. Which two techniques are appropriate for this task? (Select TWO).

Accepted Answer

Recursive Feature Elimination (RFE). Recursive Feature Elimination (RFE) (B) is a wrapper-based feature selection method that repeatedly trains a model, ranks features by importance (e.g., coefficients or feature importances), and prunes the least important ones until the desired number of features remains, directly reducing the 100-feature space to improve performance and cut training time. L1 Regularization (C) — Lasso — adds a penalty equal to the absolute value of coefficients to the loss function, which drives many feature coefficients exactly to zero, effectively performing embedded feature selection and yielding a sparse model. Normalization (A) is a scaling preprocessing step (e.g., min-max or z-score) that changes feature magnitudes but does not remove features, so it is not feature selection. One-Hot Encoding (D) is a categorical-encoding transformation that expands categorical variables into binary columns, increasing dimensionality rather than reducing it. Principal Component Analysis (E) is a dimensionality-reduction technique that creates new uncorrelated components from linear combinations of the original features, but it is not feature selection because it does not retain or select the original features.

Answer

Normalization

Answer

One-Hot Encoding

Answer

Principal Component Analysis (PCA)

Question 7

A hospital deploys an AI diagnostic assistant that analyzes medical images. The system has been in use for six months, and radiologists have reported that the AI is increasingly confident in its predictions, but sometimes misses rare conditions. The AI ethics board is concerned about overreliance and potential harm from false negatives. They want to implement a governance framework that ensures appropriate human oversight. The hospital has a limited IT budget. What is the best approach?

Accepted Answer

Implement a human-in-the-loop process where the AI flags low-confidence or rare condition predictions for mandatory radiologist review. A human-in-the-loop process that triggers mandatory radiologist review only for low-confidence or rare-condition predictions directly addresses the risk of overreliance and false negatives without overwhelming the limited IT budget. This targeted oversight ensures that the AI's increasing confidence does not lead to missed rare conditions, while still allowing routine high-confidence predictions to proceed efficiently. The approach balances safety and resource constraints by focusing human attention where the AI is most likely to err.

Answer

Add a warning to the AI interface that says 'This tool may miss rare conditions'

Answer

Require all AI predictions to be reviewed by a radiologist before final diagnosis

Answer

Increase the AI's false positive threshold to reduce missed cases

Question 8

A developer is using a pre-trained BERT model for a question-answering system. They want to ensure the model can handle out-of-vocabulary words. Which component of the BERT architecture is responsible for this?

Accepted Answer

WordPiece tokenisation. WordPiece tokenisation is the component of BERT that handles out-of-vocabulary (OOV) words by breaking them into subword units (e.g., 'playing' → 'play' + '##ing'). This allows the model to represent any word, even unseen ones, as a sequence of known subword tokens, ensuring no word is truly out of vocabulary.

Answer

Positional encoding

Answer

Feed-forward layers

Answer

Attention mechanism

Question 9

A financial institution uses an AI model to approve small business loans. The model has a high approval rate for women-owned businesses but low for minority-owned businesses. The compliance officer is concerned about disparate impact. Which governance process should be implemented first?

Accepted Answer

Conduct a bias audit and fairness assessment using relevant metrics. A bias audit and fairness assessment should be conducted first to quantify the disparate impact and identify root causes. Option A is wrong because simply removing sensitive features does not guarantee fairness and may be insufficient or illegal. Option C is wrong because publishing decision-making criteria without first addressing bias could expose the institution to liability and undermine trust. Option D is wrong because adjusting the approval threshold without thorough analysis can be arbitrary, may not address underlying bias, and could lead to reverse discrimination or mask systemic issues.

Answer

Remove gender and ethnicity features from the model

Answer

Publish the model's decision-making criteria to the public

Answer

Immediately adjust the approval threshold to equalize rates

Question 10

A company streams sensor data from IoT devices. The data arrives as JSON messages at high velocity. Which data pipeline architecture is BEST suited to handle this streaming data for near-real-time analytics?

Accepted Answer

Stream processing using Apache Kafka and Spark Streaming.. Apache Kafka acts as a distributed, fault-tolerant ingestion layer that can handle high-velocity JSON messages, while Spark Streaming processes the data in micro-batches for near-real-time analytics. This combination provides the low-latency, scalable pipeline required for streaming IoT sensor data, unlike batch or single-node approaches.

Answer

Batch processing using Hadoop MapReduce every 24 hours.

Answer

Batch processing using nightly ETL jobs.

Answer

Single-node database with periodic inserts.

Question 11

A company uses a neural network for fraud detection. The dataset has 99% legitimate, 1% fraudulent. The model achieves 99% accuracy but fails to detect most frauds. Which metric should they focus on?

Accepted Answer

Recall. Recall (sensitivity) measures the proportion of actual positives correctly identified. In this fraud detection scenario with 99% legitimate and 1% fraudulent transactions, a 99% accuracy can be achieved by simply predicting all transactions as legitimate, which yields 0% recall for the fraud class. Focusing on recall ensures the model captures the majority of fraudulent cases, addressing the critical failure to detect fraud despite high accuracy.

Answer

Precision

Answer

F1-score

Answer

AUC-ROC

Question 12

A developer is building a natural language processing system to classify customer reviews as positive, neutral, or negative. They have 50,000 labeled reviews. Which model architecture is MOST appropriate for this task?

Accepted Answer

Fine-tune a pre-trained BERT model. Fine-tuning a pre-trained BERT model is most appropriate because BERT is a transformer-based model pre-trained on a large corpus and can be fine-tuned on the 50,000 labeled reviews to achieve high accuracy with relatively little data. It captures bidirectional context, which is crucial for sentiment classification, and avoids the need for training from scratch.

Answer

Use a convolutional neural network (CNN) on raw text

Answer

Train a recurrent neural network (RNN) from scratch

Answer

Word2vec embeddings followed by logistic regression

Question 13

A company is concerned about membership inference attacks on their classification model. They have a small dataset and need to train a model that minimizes privacy leakage while maintaining high accuracy. Which technique is most appropriate?

Accepted Answer

Apply differential privacy during training. Differential privacy (DP) is the most appropriate technique because it directly addresses membership inference attacks by adding calibrated noise to the training process, mathematically bounding the model's reliance on any single data point. This ensures that an adversary cannot confidently determine whether a specific record was in the training set, which is critical for a small dataset where each sample has high influence. DP provides a formal privacy guarantee (ε-differential privacy) that balances privacy leakage against model accuracy, making it the standard defense against such attacks.

Answer

Use data augmentation to expand the dataset

Answer

Train a larger model to improve generalization

Answer

Reduce the number of training epochs

Question 14

An e-commerce company needs to update its recommendation model continuously as user preferences change. The model currently retrains from scratch every night, but the training time is too long. Which approach would reduce training time while keeping the model up-to-date?

Accepted Answer

Implement incremental learning using online gradient descent.. Incremental learning using online gradient descent updates the model parameters with each new data point or mini-batch, avoiding the need to retrain from scratch. This approach significantly reduces training time while continuously adapting to changing user preferences, making it ideal for real-time recommendation systems.

Answer

Use dimensionality reduction on features.

Answer

Switch to a simpler model.

Answer

Increase the batch size for retraining.

Question 15

A company uses a third-party LLM API to power its customer support chatbot. To prevent prompt injection attacks, which defense is MOST effective at the application layer?

Accepted Answer

Input validation and sanitization. Input validation and sanitization at the application layer is the most effective defense against prompt injection because it stops malicious instructions from ever reaching the LLM. By filtering, escaping, or rejecting inputs that contain injection patterns (e.g., 'ignore previous instructions', role-play overrides, or embedded system-prompt delimiters), the application prevents the model from being manipulated. This is a preventive control applied before inference, which is stronger than detective controls applied after the model responds.

Answer

Differential privacy during training

Answer

Rate limiting API calls

Answer

Output filtering of model responses

Question 16

A medical imaging team is developing an AI model to detect tumors from CT scans. They have 10,000 labeled scans, but the labels were created by a semi-automated process with an estimated 20% error rate (mislabeled tumor vs. no tumor). The team trains a convolutional neural network (CNN) and achieves 90% accuracy on a held-out test set that was carefully validated by an expert radiologist. However, when deployed to a new hospital's patient population, the accuracy drops to 70%. The team suspects domain shift and label noise. Which strategy is most likely to improve model robustness for the new hospital?

Accepted Answer

Use active learning to select the most uncertain predictions from the new hospital's data, then have an expert radiologist correct those labels. Active learning selects the most uncertain predictions from the new hospital's data, allowing an expert radiologist to efficiently correct the most informative labels. This directly addresses both label noise (by correcting mislabeled examples) and domain shift (by focusing on samples where the model is uncertain in the new domain). Option B is wrong because random selection may not target the most impactful errors, wasting expert effort. Option C is wrong because adding more noisy labels from the same flawed process will amplify label noise without correcting the domain shift. Option D is wrong because reducing model complexity and dropout are regularization techniques that do not fix label noise or domain shift.

Answer

Randomly select 1,000 scans from the new hospital and have them re-labeled by the radiologist

Answer

Collect 20,000 more scans with the same semi-automated labeling process

Answer

Reduce the CNN's number of layers and apply dropout to combat overfitting

Question 17

Which stage of the AI project lifecycle involves splitting data into training, validation, and test sets?

Accepted Answer

Data preparation. Data preparation is the stage where raw data is cleaned, transformed, and split into training, validation, and test sets. This split is a core part of preparing data for model training — the training set teaches the model, the validation set tunes hyperparameters, and the test set provides an unbiased final evaluation. It occurs before model training and evaluation, making data preparation the correct stage.

Answer

Model evaluation

Answer

Data acquisition

Answer

Problem definition

Question 18

Refer to the exhibit. A security auditor identifies a critical vulnerability that could allow an attacker to manipulate model inputs to cause misclassification. Which configuration setting is most directly responsible for this vulnerability?

Accepted Answer

enable_adversarial_defense = false. The vulnerability described is an adversarial attack on model inputs, which directly exploits the absence of adversarial defenses. Setting `enable_adversarial_defense = false` disables mechanisms like adversarial training or input perturbation detection that prevent misclassification from manipulated inputs. This configuration is the most direct root cause because it explicitly turns off the defense designed to counter such attacks.

Answer

enable_input_sanitization = true

Answer

audit_level = basic

Answer

pii_detection = enabled

Question 19

A generative AI model produces images from text prompts. The outputs are often blurry and lack fine details. Which model type is MOST likely being used, and which improvement would best address this issue?

Accepted Answer

Variational Autoencoder (VAE); switch to a diffusion model. Variational Autoencoders (VAEs) are known for producing blurry outputs because their loss function (ELBO) encourages pixel-wise averaging, which smooths out fine details. Diffusion models, by contrast, iteratively denoise a random field, learning to reconstruct high-frequency details through a multi-step reverse process, directly addressing the blurriness issue.

Answer

Variational Autoencoder (VAE); switch to a Generative Adversarial Network (GAN)

Answer

Generative Adversarial Network (GAN); increase the discriminator's capacity

Answer

Diffusion model; use a larger batch size during training

Question 20

Which privacy-preserving technique allows a model to be trained across decentralized data sources without the raw data ever leaving each source?

Accepted Answer

Federated learning. Federated learning trains models locally on each device or server and only shares model updates, preserving data locality.

Answer

Homomorphic encryption

Answer

Secure multi-party computation

Answer

Differential privacy