Amazon AIF-C01 Online Practice Questions and Exam Preparation

Amazon AIF-C01 Online Questions & Answers

• Question 161:

  A company wants to use a large language model (LLM) to develop a conversational agent. The company needs to prevent the LLM from being manipulated with common prompt engineering techniques to perform undesirable actions or expose sensitive information. Which action will reduce these risks?

  A. Create a prompt template that teaches the LLM to detect attack patterns.
  B. Increase the temperature parameter on invocation requests to the LLM.
  C. Avoid using LLMs that are not listed in Amazon SageMaker.
  D. Decrease the number of input tokens on invocations of the LLM.
  A. Create a prompt template that teaches the LLM to detect attack patterns.

  ---

  Creating a prompt template that teaches the LLM to detect attack patterns is the most effective way to reduce the risk of the model being manipulated through prompt engineering.

• Question 162:

  An AI practitioner needs to improve the accuracy of a natural language generation model. The model uses rapidly changing inventory data. Which technique will improve the model's accuracy?

  A. Transfer learning
  B. Federated learning
  C. Retrieval Augmented Generation (RAG)
  D. One-shot prompting
  C. Retrieval Augmented Generation (RAG)

  ---

  RAG enhances a language model by fetching up-to-date, domain-specific data (e.g., current inventory) at inference time and conditioning the generation on those facts, ensuring the output reflects the latest information and improving accuracy without retraining the core model.

• Question 163:

  A company wants to identify harmful language in the comments section of social media posts by using an ML model. The company will not use labeled data to train the model. Which strategy should the company use to identify harmful language?

  A. Use Amazon Rekognition moderation.
  B. Use Amazon Comprehend toxicity detection.
  C. Use Amazon SageMaker AI built-in algorithms to train the model.
  D. Use Amazon Polly to monitor comments.
  B. Use Amazon Comprehend toxicity detection.

  ---

  Amazon Comprehend provides pre-trained NLP models, including toxicity detection, to analyze text for harmful language. Since the company does not plan to use labeled data for training, Amazon Comprehend is a suitable choice because it does not require custom training and can automatically detect toxic or harmful content in comments.

• Question 164:

  A company is developing an ML model to make loan approvals. The company must implement a solution to detect bias in the model. The company must also be able to explain the model's predictions. Which solution will meet these requirements?

  A. Amazon SageMaker Clarify
  B. Amazon SageMaker Data Wrangler
  C. Amazon SageMaker Model Cards
  D. AWS AI Service Cards
  A. Amazon SageMaker Clarify

  ---

  Amazon SageMaker Clarify is a tool that helps detect bias in machine learning models and provides explainability for model predictions. It allows users to understand the factors influencing a model's predictions and assess whether the model is fair across different demographic groups. This is exactly what the company needs for detecting bias and explaining model decisions, especially for high-stakes applications like loan approvals.

  Bias detection: SageMaker Clarify can analyze the training data and the model's predictions to identify and mitigate bias.

  Explainability: It provides features for explaining the predictions made by the model, which helps users understand the reasons behind the model's decisions.

• Question 165:

  A company is building a conversational AI assistant. The company plans to use Amazon Bedrock Guardrails to prevent users from attempting to bypass the AI assistant's built-in safety capabilities.

  Which type of Bedrock Guardrail policy will meet this requirement?

  A. Contextual grounding check
  B. Denied topics
  C. Prompt attacks
  D. Sensitive information filters
  C. Prompt attacks

  ---

  Explanation

  Prompt attack policies in Amazon Bedrock Guardrails are designed to detect and block attempts to manipulate the model into ignoring or bypassing its built-in safety instructions.

• Question 166:

  HOTSPOT

  An AI practitioner is using an Amazon Nova text model. The AI practitioner wants to apply prompt engineering techniques to ensure that prompts support an iterative refinement process.

  Select the correct prompt design component from the following list for each definition. Select each prompt design component one time.

  

  

• Question 167:

  Which option is a benefit of using infrastructure as code (IaC) in machine learning operations (MLOps)?

  A. IaC eliminates the need for hyperparameter tuning.
  B. IaC always provisions powerful compute instances, contributing to the training of more accurate models.
  C. IaC streamlines the deployment of scalable and consistent ML workloads in cloud environments.
  D. IaC minimizes overall expenses by deploying only low-cost instances.
  C. IaC streamlines the deployment of scalable and consistent ML workloads in cloud environments.

  ---

  By describing infrastructure in code, teams can version, automate, and repeatably provision resources, ensuring that ML training and inference environments are consistent, scalable, and easy to reproduce across development, testing, and production.

• Question 168:

  A company is developing a generative AI application to automatically generate product descriptions for an ecommerce website. The product descriptions must consist of paragraphs of text that are consistent in style and tone. The application must generate thousands of unique descriptions each day. Which type of generative model will meet these requirements?

  A. A variational autoencoder (VAE) model
  B. A transformer-based model
  C. A diffusion model
  D. A generative adversarial network (GAN) model
  B. A transformer-based model

  ---

  Transformer-based models, such as large language models, are designed to generate coherent and contextually consistent text at scale. They are well-suited for creating thousands of unique, styled product descriptions daily.

• Question 169:

  A company has developed a generative AI model for customer segmentation. The model has been deployed in the company's production environment for a long time. The company recently noticed some inconsistency in the model's responses. The company wants to evaluate model bias and drift. Which AWS service or feature meets these requirements?

  A. Amazon SageMaker Model Monitor
  B. Amazon SageMaker Clarify
  C. Amazon SageMaker Model Cards
  D. Amazon SageMaker Feature Store
  A. Amazon SageMaker Model Monitor

  ---

  SageMaker Model Monitor continuously collects inference and ground-truth data to automatically detect data drift, concept drift, and bias shifts in production models. By configuring baseline metrics and thresholds, the company can track inconsistencies in responses over time and receive alerts whenever bias or drift exceed acceptable limits.

• Question 170:

  A company has documents that are missing some words because of a database error. The company wants to build an ML model that can suggest potential words to fill in the missing text.

  Which type of model meets this requirement?

  A. Topic modeling
  B. Clustering models
  C. Prescriptive ML models
  D. BERT-based models
  D. BERT-based models

  ---

  BERT-based models (Bidirectional Encoder Representations from Transformers) are suitable for tasks that involve understanding the context of words in a sentence and suggesting missing words. These models use bidirectional training, which considers the context from both directions (left and right of the missing word) to predict the appropriate word to fill in the gaps.