AIF-C01 practice questions 61–70

1. Question 61 · 417 in bank

   An AI practitioner wants to use a foundation model (FM) to design a search application. The search application must handle queries that have text and images. Which type of FM should the AI practitioner use to power the search application?

   • AMulti-modal embedding model
   • BText embedding model
   • CMulti-modal generation model
   • DImage generation model

   Show answer

   Answer: A

   Explanation

   The answer is A. Multi-modal embedding model. A multi-modal embedding model is a type of foundation model that can process and understand both text and images. This makes it suitable for powering a search application that handles queries containing both text and images. Here's a breakdown of the other options: B. Text embedding model: This type of model is only designed to process text data, so it wouldn't be suitable for handling image queries. C. Multi-modal generation model: This type of model is designed to generate text or images, not to search for them. D. Image generation model: This type of model is only designed to generate images, not to search for them.https://www.examtopics.com/exams/amazon/aws-certified-ai-practitioner-aif-c01/view/5/#

   Source: Adapted from community discussion — Moon

   Next question

2. Question 62 · 417 in bank

   A company uses a foundation model (FM) from Amazon Bedrock for an AI search tool. The company wants to fine-tune the model to be more accurate by using the company's data. Which strategy will successfully fine-tune the model?

   • AProvide labeled data with the prompt field and the completion field.
   • BPrepare the training dataset by creating a .txt file that contains multiple lines in .csv format.
   • CPurchase Provisioned Throughput for Amazon Bedrock.
   • DTrain the model on journals and textbooks.

   Show answer

   Answer: A

   Explanation

   Fine-tuning and provisioned throughput are two separate processes in Amazon Bedrock: Fine-tuning is the process of training the model on specific labeled data to improve its accuracy for particular tasks. This can be done without purchasing provisioned throughput.Provisioned throughput is required after fine-tuning, specifically for using the custom model for inference. It's not needed for the fine-tuning process itself.To test and deploy your model, you need to purchase Provisioned Throughput.

   Source: Adapted from community discussion — Dandelion2025

   Next question

3. Question 63 · 417 in bank

   A law firm wants to build an AI application by using large language models (LLMs). The application will read legal documents and extract key points from the documents. Which solution meets these requirements?

   • ABuild an automatic named entity recognition system.
   • BCreate a recommendation engine.
   • CDevelop a summarization chatbot.
   • DDevelop a multi-language translation system.

   Show answer

   Answer: C

   Explanation

   C. Develop a summarization chatbot. Explanation: A summarization chatbot can leverage large language models (LLMs) to automatically read and extract key points from legal documents by summarizing the content. This approach aligns well with the firm's need to condense lengthy documents into concise, relevant summaries, making it easier for users to quickly understand the main points without reading the entire document. LLMs are highly effective at summarization tasks, especially when fine-tuned on domain-specific data like legal text.

   Source: Adapted from community discussion — jove

   Next question

4. Question 64 · 417 in bank

   A company wants to use AI to protect its application from threats. The AI solution needs to check if an IP address is from a suspicious source. Which solution meets these requirements?

   • ABuild a speech recognition system.
   • BCreate a natural language processing (NLP) named entity recognition system.
   • CDevelop an anomaly detection system.
   • DCreate a fraud forecasting system.

   Show answer

   Answer: C

   Explanation

   An anomaly detection system can analyze patterns and behaviors, such as IP address access patterns, to detect any deviations from the norm, which could indicate suspicious or malicious activity. An anomaly detection model can flag unusual access attempts, such as those from suspicious IP addresses, making it well-suited for threat detection. Fraud forecasting (option D) typically focuses on predicting potential fraud patterns rather than real-time anomaly detection, so it would not directly address the need to check IP addresses for suspicious activity. Thus, option C is the most suitable choice for identifying suspicious IP addresses in this scenario.

   Source: Adapted from community discussion — Udyan

   Next question

5. Question 65 · 417 in bank

   Which feature of Amazon OpenSearch Service gives companies the ability to build vector database applications?

   • AIntegration with Amazon S3 for object storage
   • BSupport for geospatial indexing and queries
   • CScalable index management and nearest neighbor search capability
   • DAbility to perform real-time analysis on streaming data

   Show answer

   Answer: C

   Explanation

   Scalable index management and nearest neighbor search capability: Amazon OpenSearch Service provides built-in support for vector search, which allows for efficient nearest neighbor search (such as k-nearest neighbors or k-NN) in large datasets. This is essential for vector databases, which store high-dimensional data (such as embeddings from machine learning models) and support fast similarity search. The scalable index management ensures that these searches can be performed efficiently even with large datasets.

   Source: Adapted from community discussion — Jessiii

   Next question

6. Question 66 · 417 in bank

   Which option is a use case for generative AI models?

   • AImproving network security by using intrusion detection systems
   • BCreating photorealistic images from text descriptions for digital marketing
   • CEnhancing database performance by using optimized indexing
   • DAnalyzing financial data to forecast stock market trends

   Show answer

   Answer: B

   Explanation

   The correct answer is B. Creating photorealistic images from text descriptions for digital marketing. Generative AI models are designed to create new content, such as text, images, audio, or code. Creating images from text descriptions is a prime example of this capability. Here's why the other options are not primarily use cases for generative AI: A. Improving network security by using intrusion detection systems: While AI can be used for intrusion detection, this is more of a discriminative or predictive task (classifying network traffic as malicious or benign), not generating new content. C. Enhancing database performance by using optimized indexing: This is related to database management and optimization, not content generation. D. Analyzing financial data to forecast stock market trends: This involves statistical analysis and prediction based on existing data, again a predictive task, not generating new content.

   Source: Adapted from community discussion — Moon

   Next question

7. Question 67 · 417 in bank

   A company wants to build a generative AI application by using Amazon Bedrock and needs to choose a foundation model (FM). The company wants to know how much information can fit into one prompt. Which consideration will inform the company's decision?

   • ATemperature
   • BContext window
   • CBatch size
   • DModel size

   Show answer

   Answer: B

   Explanation

   A company needs to know the maximum input size for a single prompt when choosing a Foundation Model (FM) in Amazon Bedrock. A. Temperature: This controls the randomness of the output, not the input prompt length. Temperature affects creativity, not input size. B. Context window: This defines the maximum length of the input prompt the model can process. It directly limits how much information can be included. C. Batch size: This is the number of prompts processed at once, affecting throughput, not individual prompt length. It's about processing multiple prompts efficiently. D. Model size: This relates to the model's overall capacity and complexity, not directly to the input prompt length. Size impacts performance, not input limits. Therefore, B. Context window is the correct answer.

   Source: Adapted from community discussion — Moon

   Next question

8. Question 68 · 417 in bank

   A company wants to make a chatbot to help customers. The chatbot will help solve technical problems without human intervention. The company chose a foundation model (FM) for the chatbot. The chatbot needs to produce responses that adhere to company tone. Which solution meets these requirements?

   • ASet a low limit on the number of tokens the FM can produce.
   • BUse batch inferencing to process detailed responses.
   • CExperiment and refine the prompt until the FM produces the desired responses.
   • DDefine a higher number for the temperature parameter.

   Show answer

   Answer: C

   Next question

9. Question 69 · 417 in bank

   A company wants to use a large language model (LLM) on Amazon Bedrock for sentiment analysis. The company wants to classify the sentiment of text passages as positive or negative. Which prompt engineering strategy meets these requirements?

   • AProvide examples of text passages with corresponding positive or negative labels in the prompt followed by the new text passage to be classified.
   • BProvide a detailed explanation of sentiment analysis and how LLMs work in the prompt.
   • CProvide the new text passage to be classified without any additional context or examples.
   • DProvide the new text passage with a few examples of unrelated tasks, such as text summarization or question answering.

   Show answer

   Answer: A

   Explanation

   A: Provide examples of text passages with corresponding positive or negative labels in the prompt followed by the new text passage to be classified. Explanation: This strategy is known as few-shot prompting, where the prompt includes a few examples of labeled data (text passages with positive or negative sentiment) before asking the model to classify the new text passage. This helps the large language model (LLM) understand the task and align its output with the desired format. Why not the other options? B: Provide a detailed explanation of sentiment analysis and how LLMs work in the prompt: Explaining the concept of sentiment analysis is unnecessary for the model, as it does not improve the model's ability to classify text. C: Provide the new text passage to be classified without any additional context or examples: Without examples, the LLM might not correctly infer the task or format of the output, leading to inconsistent or incorrect results.

   Source: Adapted from community discussion — Moon

   Next question

10. Question 70 · 417 in bank

    A security company is using Amazon Bedrock to run foundation models (FMs). The company wants to ensure that only authorized users invoke the models. The company needs to identify any unauthorized access attempts to set appropriate AWS Identity and Access Management (IAM) policies and roles for future iterations of the FMs. Which AWS service should the company use to identify unauthorized users that are trying to access Amazon Bedrock?

    • AAWS Audit Manager
    • BAWS CloudTrail
    • CAmazon Fraud Detector
    • DAWS Trusted Advisor

    Show answer

    Answer: B

    Explanation

    AWS CloudTrail: CloudTrail records all API requests made to AWS services, including Amazon Bedrock. By using CloudTrail, the security company can track and log all access attempts, including any unauthorized attempts, to access Amazon Bedrock. This helps the company identify and respond to unauthorized access, and also provides detailed logs for setting up appropriate IAM policies and roles.

    Source: Adapted from community discussion — Jessiii

    Next question

Keyboard:J Kmove between questionsA–Dselect an answerRshow or hide the answer← →change page

AIF-C01 practice, question 61. 417 questions available.