AWS AIF-C01 Exam Prep

All Questions

Browse exam-style questions one by one from the list. Filter the ones you missed and quickly find the questions you want to review.

G1An ML team wants to build and run a series of machine learning workflow steps, such as data preparation, training, evaluation, and deployment, as an automated and reproducible pipeline. Which SageMaker feature is the MOST suitable?
G1A hospital wants to build a model that uses labeled X-ray images previously diagnosed by radiologists to determine whether a new chest X-ray image shows a lesion or no lesion. There was also a proposal to estimate the size of the lesion as a number, but the only requirement this time is the lesion or no-lesion determination. Which type of task is this?
G1An ML team is dealing with a model whose accuracy on the training data keeps improving as training continues, while its accuracy on the validation data starts to worsen partway through. Which is the MOST appropriate measure to curb this overfitting?
G1An ML team validated a demand forecasting model it had built and found that it achieves very high accuracy on the training data but its accuracy drops significantly on new, unseen data. What is this state called?
G1An ML team has finished collecting data and is planning the work for the data preparation phase before moving into model training. While checking the mapping between each lifecycle phase and its work, which tasks are appropriate for this phase? (Choose TWO.)
G2A legal team wants to feed an entire long contract to a model as a single input and have it summarized. What is the name of the property that represents how many tokens a model can take as input and produce as output at one time?
G2At a company executive meeting, the team is gathering material to decide whether to continue a generative AI project. What is the name of the financial metric that shows how much profit or value was gained relative to the cost invested in the project?
G2A developer is comparing inference parameters that finely control output diversity, while being aware of the difference from top-k. What is the name of the setting that controls diversity by narrowing the output candidates to the top words until a cumulative probability reaches a set value?
G2A development team is estimating the cost of on-demand usage of Amazon Bedrock. The pricing lists a unit price based on the amount of input and output, and the upper limit on the input a model can accept at one time (the context window) is also expressed in the same unit. Which term corresponds to this unit?
G2A company is designing an operational structure for a foundation model. In mapping each stage of the lifecycle to a responsible team, which stages are involved after the model begins to be used in production (from deployment onward)? (Choose TWO.)
G3A developer of an image generation application is organizing ways to control the output. In image and text generation, what is the name of the prompt specification that explicitly states elements to suppress from the output, such as specific words or styles, so they are not included?
G3An AI team feels limited by evaluating heavily paraphrased generated text using only surface-match metrics. Which metric evaluates the quality of generated text not only by surface-level word matching but also by semantic closeness using embeddings?
G3In an AI team's design review, there is debate over whether to use RAG or fine-tuning. To share decision criteria, which option BEST explains how to choose between the two?
G3A company performed fine-tuning but did not achieve the expected results, and decided to review its approach starting from data preparation. Which is the MOST appropriate data preparation for improving the effectiveness of fine-tuning?
G3A company is checking how far Bedrock Guardrails can meet the safety requirements for a chatbot. Which protections (filters) can be configured with Amazon Bedrock Guardrails? (Choose TWO.)
G4A company wants to add a mechanism to automatically suppress outputs that contain harmful expressions or undesirable topics before publicly launching a chatbot built with Bedrock. Which feature is MOST appropriate for this requirement?
G4A healthcare company is formulating an AI system adoption policy. The company places a high priority on protecting the personal data of patients handled by AI and preventing unauthorized access and data breaches. Which characteristic of responsible AI BEST describes this consideration?
G4A company wants to ensure it can respond quickly and appropriately when issues such as harmful outputs or critical errors occur in a production AI system. What should the company prepare in advance to meet this requirement?
G4A company is defining the fairness review procedure for a credit scoring model. The company wants to examine whether there are differences in accuracy or error rates across demographic groups such as age group, gender, and region. What is the name of this type of analysis?
G4A company is formulating an AI governance policy and wants to distinguish between the core dimensions of responsible AI identified by AWS and general system performance requirements. Which TWO of the following are characteristics of responsible AI? (Choose TWO.)
G5A company wants to design access permissions for each component of an AI application that uses Bedrock, allowing only the operations required for business and granting no additional permissions. Which principle describes this approach?
G5During a design review at a company, audit requirements and data storage location requirements were being confused. Which statement BEST describes the relationship between enabling data logging (recording) and the physical storage location of data (region)?
G5A company is operating a generative AI application on Amazon Bedrock and is organizing the division of responsibilities with AWS for an upcoming security audit. Under the AWS shared responsibility model, which area is MOST appropriately handled by AWS?
G5A company wants to collect metrics on the usage of Amazon Bedrock (such as the number of model invocations and the number of input and output tokens) and monitor them with dashboards and alarms. Which AWS service is MOST suitable for this requirement?
G5A company is building the operational framework for an AI data infrastructure. Which TWO of the following are activities that are routinely performed to maintain healthy data governance, while being distinct from model development or infrastructure operations tasks? (Choose TWO.)
G6An ML team is organizing the metrics to include in a classification model's evaluation report. What is the name of the most basic metric, used to evaluate a classification model, that represents the proportion of predictions that were classified correctly out of all predictions?
G6An ML team is analyzing the breakdown of a classification model's errors. What is the name of the tool used to organize prediction results into a table of the combinations of actual class and predicted class (true positive, false positive, true negative, false negative) so that error tendencies can be understood?
G6In an ML organization, teams are wastefully re-creating the same features separately. Which SageMaker feature lets you store features (feature data) used by multiple teams and models centrally so they can be shared and reused?
G6A company's application emits event logs whose set of fields changes from request to request. The logs are in JSON format that expresses hierarchy with tags or keys but has no strict fixed schema per column. Which type of data is this?
G6An ML team is reviewing remediations for a model that has overfit. As it reconsiders how training is run and how data is handled, which are effective measures for reducing overfitting? (Choose TWO.)
G7An AI team is organizing a table of when each model selection criterion should be prioritized. In which situation is it MOST appropriate to prioritize task fit (accuracy for the use case)?
G7Amazon Bedrock provides foundation models from multiple vendors (Anthropic, Amazon, Meta, and others) through a single API. Which is the MOST appropriate benefit of this?
G7At a company's planning meeting, the benefits of adopting generative AI are being organized. When explaining the benefits of generative AI, which is the MOST appropriate meaning of adaptability?
G7A company is examining the barriers to adopting generative AI. Which is the MOST appropriate reason that generative AI (especially foundation models from managed services) is considered relatively easy to start using?
G7An AI team is building a checklist for selecting a foundation model. As specified in the question, performance and cost have already been evaluated. Which factors should be considered beyond performance and cost? (Choose TWO.)
G8An AI practitioner wants to limit the candidates for the next token to only the top few with the highest probability before selection, suppressing extreme outlier words. Which inference parameter performs this control?
G8A company is strengthening hallucination defenses for a RAG chatbot. Which Bedrock guardrails mechanism verifies whether a RAG-generated response is actually grounded in the content of the provided reference documents (sources), that is, whether it strays from the evidence?
G8A developer is organizing the role of each component of a prompt. What is the name of the component that, like 'Human:' or 'Assistant:', indicates where the model's response begins and its format, that is, the start position and format of the output?
G8A company wants to build, with as little effort as possible, a RAG application on Amazon Bedrock that ingests internal documents and answers based on their content. Which Bedrock feature is the BEST fit?
G8An AI team is choosing metrics to automatically evaluate the quality of summarization and translation. Which mappings of automatic text-generation evaluation metrics to their representative target tasks are correct? (Choose TWO.)
G9A company is deciding on a format to quantitatively report the results of a fairness review of a credit model to executives and auditors. What is the collective name for the metrics used to numerically measure whether a model's predictions are fair across groups?
G9A company is comparing a high-accuracy complex model with a simple, easy-to-explain model when selecting a credit model. Which is the MOST appropriate description of the general relationship between a model's performance (accuracy) and interpretability (ease of explanation)?
G9In a company review, a model trained on biased historical hiring data was found to reproduce the same bias. What is the name of the problem in which bias contained in the training data appears directly as bias in the model's predictions?
G9In a company's risk review, the team is evaluating the impact of delivering generative AI output that contains information contrary to fact (a hallucination) as is. Which is the MOST appropriate risk that could arise for end users?
G9A financial institution is selecting a model for a process where audits require explanations of prediction reasoning. The policy is to prioritize ease of explanation over accuracy. Which models are generally considered to have high interpretability (ease of explanation)? (Choose TWO.)
G10A company wants to use data containing personally identifiable information for analysis and ML training while protecting privacy. Which option represents the effort to achieve this through processing such as anonymization and masking?
G10A healthcare company is required by regulation to keep patient data stored within a specific country. Which is the MOST appropriate way to address this 'data storage location' requirement on AWS?
G10A company is looking for an approach to drive AI adoption across the entire organization. Which AWS framework helps an organization that uses AI systematically advance its transformation from perspectives such as business, people, governance, and operations?
G10Which AWS framework organizes generative AI usage patterns into tiers (scopes) based on how involved your organization is (from using off-the-shelf services to building your own model) and examines the security risks of each?
G10A company is researching external frameworks it can reference when establishing AI governance policy. Which of the following are appropriate as frameworks or regulations referenced in relation to AI and data governance? (Choose TWO.)
G11A company wants to deploy a trained model for production use as low-latency real-time inference while minimizing the overhead of building and operating servers themselves. Which deployment approach BEST meets this requirement?
G11A retail company has accumulated five years of monthly sales data. The data is arranged in chronological order, contains seasonal cycles and trends, and the company wants to use this ordering to predict future values. Which data type BEST describes this data?
G11An ML team wants to compare models before a decision threshold has been set. What is the name of the metric that represents the area under the ROC curve and expresses the discrimination performance of a binary classifier in a single value that is independent of the threshold?
G11An AI practitioner has completed model training but accuracy has not reached the target. The practitioner wants to improve model performance by adjusting settings such as the learning rate and batch size. Which process does this describe?
G11At a training session, it was noted that image-related AWS AI services are often confused. Which TWO of the following correctly pair the use cases of 'image analysis' and 'document text extraction (OCR)' with the corresponding AWS services? (Choose TWO.)
G12A company is planning to improve its FAQ search. The plan is to convert each text to a representation that can numerically compute semantic similarity and store it, so that even when a user's question and an FAQ use different wording, a search will return a match if the meanings are close. Which term describes this representation?
G12A design company wants to introduce generative AI that generates new images matching a text description as input, in order to streamline the creation of advertising materials. Which type of model is MOST appropriate for this use case?
G12A company is building a configuration to ingest a long internal manual into an Amazon Bedrock knowledge base and search only the relevant sections to use in generating answers. Prior to ingestion, the document must be split into meaningful units suitable for embedding conversion and search. Which term describes this splitting process?
G12A development team wants to build a generative AI application by calling high-performance foundation models from multiple vendors through a single API without managing infrastructure, and also needs the flexibility to switch between models depending on the use case. Which AWS service is MOST suitable?
G12A company is considering where to apply generative AI (foundation models) within its own services, while distinguishing these cases from situations where traditional approaches are sufficient. Which TWO of the following are use cases where generative AI (foundation models) can be effectively applied? (Choose TWO.)
G13A company is implementing safety measures for a publicly available chatbot using Bedrock features. Which feature filters and blocks harmful content, disallowed topics, and confidential information at both input and output?
G13During a training session, a participant is learning how to adjust the behavior of generative AI. Which of the following BEST describes the role that inference parameters (such as temperature, Top-P, and maximum token count) play in the responses of generative AI?
G13A developer is looking for ways to address a task where output format is unstable. The developer wants to provide several input and output examples within the prompt to a foundation model and then run the actual task, stabilizing the accuracy and format of the output. What is this technique called?
G13A company is integrating generative AI into a customer support chat interface. It is important that the response starts returning immediately after the user submits a message, without making users wait. Which criterion should be given the MOST weight when selecting a model?
G13A company is preparing to fine-tune a foundation model for internal use. Which TWO of the following represent appropriate data governance perspectives to follow regarding the data to be used for training? Principles take priority over efficiency. (Choose TWO.)
G14A media company is considering a project to produce content using the voices and likenesses of real people with generative AI. When reviewing the project from a responsible AI perspective before release, which concern should receive particular attention?
G14A company is taking stock of services that can be used for bias detection in order to set up a responsible AI review process. Which of the following services CANNOT be used for the purpose of detecting and evaluating bias hidden in training data or model predictions?
G14As a pre-release review of a recruiting support model, a company wants to detect whether the training data or the model's predictions are biased toward specific demographic groups. A mechanism for monitoring drift after deployment is already in place separately. Which AWS service is the MOST suitable?
G14For a loan-screening AI, an officer wants to be able to understand on an individual basis why a particular application was decided to be rejected. Which characteristic of responsible AI BEST fits this requirement?
G14For an ML model in production, a company wants to work on monitoring performance degradation after deployment and on documenting the model's intended use and risks for a governance audit. Which TWO AWS features help with these? (Choose TWO.)
G15A company is strengthening its governance for generative AI. What is the name of the recently emphasized concept of recording the usage history of generative AI (when, who, and how a model was invoked) so that it can be audited later?
G15A company is establishing a data handling policy. What is the data governance practice of dividing data by sensitivity (public, internal, confidential, and so on), labeling it, and changing the handling rules accordingly called?
G15In a generative AI application, a company wants to put countermeasures in place against 'prompt injection,' an attack that tries to hijack the model's intended behavior with instructions slipped into user input. Which of the following is the MOST appropriate countermeasure?
G15For an AI application that uses Bedrock, a company wants to control in fine-grained detail which users and applications can access it. Which AWS feature is the MOST appropriate?
G15A company is concerned that the AI governance it established may become a hollow formality over time. Which TWO of the following are appropriate operational elements for keeping AI governance functioning continuously rather than as a one-time effort? (Choose TWO.)
G16An ML team is defining quality control rules for model development. In machine learning, which of the following BEST describes the main purpose of splitting data into training (train), validation, and test sets?
G16A payment services company wants to introduce a mechanism that detects fraud (such as impersonation and fraudulent orders) in online transactions and new registrations in real time, without ML experts. Which managed AWS service is the MOST suitable?
G16A business analyst wants to build demand forecasting in-house without the help of an engineer. Which AWS service lets you build and run machine learning prediction models with only GUI operations, without writing code (including time-series forecasting)?
G16A financial institution is planning to automatically route inquiry emails arriving at its support desk. Three years of past emails all retain a 'spam / not spam' decision assigned by staff, and the company will use this data to build a model that automatically classifies new emails. Which learning approach is the MOST suitable for this requirement?
G16An AI practitioner is trying to choose an inference method (real-time / asynchronous / batch, and so on) suited to a use case. Which TWO of the following are appropriate considerations for choosing a method? (Choose TWO.)
G17An employee wants to improve the answer quality of an internal FAQ bot. The plan is to improve output quality by refining how instructions and example sentences are written, without retraining the model or adding an external database. Which concept represents this approach?
G17A development team wants to have generative AI write the corresponding source code from comments written in natural language (what they want to do). Which generative AI use case does this purpose fall under?
G17A company is trying to adapt a pretrained foundation model that can only give general-purpose responses, for its own specific task, by additionally training it with labeled data. Which lifecycle stage does this work correspond to?
G17A company is planning training to give non-engineer employees hands-on experience with generative AI. Which AWS service is a learning- and experiment-oriented 'playground' that lets you easily prototype and share generative AI apps in a browser, without writing code?
G17A company is trying to accurately state not only 'what it can do' but also 'what it cannot do' in a generative AI adoption proposal. Which TWO of the following are appropriate as inherent limitations of generative AI? (Choose TWO.)
G18A company wants to apply the knowledge of a model trained on one task to a related but different task, to build a new model efficiently with little data. The aim is to reduce the cost of data collection. Which technique represents this idea?
G18A developer wants to improve the output quality of a generative AI app they just built. They want to use the easiest-to-try customization technique that improves output by refining how instructions and examples are written, without changing the model's weights. Which technique is the MOST suitable?
G18A security team is classifying attacks on AI systems by the stage at which they occur. What is the attack called in which an attacker injects malicious data into the training data to intentionally distort the model's behavior or degrade its performance?
G18A developer is delegating a classification task to a foundation model and wants to try doing it without incurring the cost of preparing examples. What is the prompt technique called in which you give only the description of the task you want done, without showing any examples, and have it run?
G18An AI practitioner is comparing foundation models on Amazon Bedrock for an internal FAQ chatbot (Japanese only, text only, no additional model tuning needed). The speed at which a response starts coming back drives the experience, and because a large volume of questions is processed continuously, suppressing the bill is also necessary. Which TWO selection criteria directly correspond to these requirements? (Choose TWO.)
G19A company is defining operational rules for using generative AI in high-impact decisions such as medical diagnosis and credit assessment. After weighing the balance between efficiency and accountability, which practice is the MOST appropriate from a responsible AI perspective?
G19During a company's legal review, the team is organizing the legal risks a company could face when generative AI produces content that closely resembles an existing copyrighted work. Which legal risk is the MOST appropriate for this case?
G19A team places importance on having documented information about an AI model it uses, such as what data it was trained on, its intended uses, and its known limitations. Which responsible AI characteristic BEST matches this property?
G19A company wants to detect "drift", where the distribution of input data gradually shifts from training time, for a machine learning model deployed to production, and to catch quality degradation early. Which AWS capability is the MOST suitable?
G19A company is classifying risks by their nature in order to establish an internal policy for using generative AI. Which of the following are appropriate as legal and ethical risks associated with the use of generative AI? (Choose TWO.)
G20A company is planning an AI service for Europe and is researching the regulations that apply. Which regulation strictly governs the handling of personal data within the EU, including the data subject's consent, rights, and cross-border transfers?
G20A company wants to continuously record and automatically evaluate whether the configuration of its AWS resources, including AI workloads, complies with standards defined by internal rules and regulations. Which AWS service is the MOST suitable?
G20A company wants to encrypt the data it stores in Amazon S3 for an AI application and to centrally create, manage, and rotate the keys used for that encryption. Which AWS service is the MOST suitable?
G20A company wants to let users verify the trustworthiness of a generative AI's answers. What is the practice of explicitly showing the references (citations) and data sources to indicate which materials an answer is based on called?
G20A company runs a generative AI application on Amazon Bedrock and is confirming the division of work with AWS as it develops a security plan. Under the shared responsibility model, which are the responsibilities of this company (the customer)? (Choose TWO.)
G21In an ML project, the team is processing cleansed raw data to design new input variables that help prediction (for example, deriving "day of week" or a "start-of-month flag" from a date). Which step does this work correspond to?
G21A company wants to build a desk that automatically responds to frequently asked questions. Which managed AWS service is used to build a conversational interface (chatbot or voice bot) using voice and text?
G21An ML team is trying to ensure the reliability of the evaluation figure used in the model's final report. Which dataset is held back until the end, used neither for training nor for hyperparameter tuning, and used to evaluate the model's final generalization performance?
G21At a company study session, the difference between machine learning and traditional programming came up as a topic. Compared with traditional procedural programs in which people write out all the processing, which description BEST represents machine learning (ML)?
G21An ML team wants to reduce the effort of preparing training data and building models. Which SageMaker capabilities support the work of labeling training data and automatically building a model from data without writing code? (Choose TWO.)
G22A company is considering whether a single foundation model can be shared across multiple business units. What is the property of a foundation model that functions broadly across multiple domains, such as healthcare, law, and general conversation, rather than being limited to a specific field, called?
G22A company is deciding where to use generative AI in its operations and where not to. Among the following situations, which is the MOST appropriate as a situation where using generative AI is not suitable (where caution is warranted)?
G22A company is trying to decide, based on cost-effectiveness, whether to keep using a fine-tuned custom model on Bedrock. Which is the MOST appropriate cost trade-off of using a custom model?
G22A company is planning quality checks along the foundation model lifecycle. Before deploying to production, which stage measures the model's performance and quality with test data or benchmarks to confirm whether it is good enough?
G22A company is comparing methods for using a general-purpose foundation model adapted to its own operations. Distinguishing them from infrastructure optimization, which are representative methods for adapting a pretrained foundation model to the company's specific use case? (Choose TWO.)
G23In a company's risk assessment, the team is classifying risks specific to generative AI. What is the risk called where confidential information a user enters in a prompt, or sensitive data a model has learned, is unintentionally leaked to the outside through output or logs?
G23A research team wants to summarize multiple research papers spanning dozens of pages together in a single request. Without splitting the documents, which is the MOST important model factor that determines whether a long input can be passed at once?
G23A company wants to improve the inconsistency in how its model's responses follow instructions. Which method trains the model on data made of "instruction (prompt) and desired response" pairs to raise its ability to follow instructions?
G23An AI team is explaining the reason for adopting RAG to executives. Without an additional mechanism, a foundation model (FM) can sometimes fail to answer accurately about the company's latest documents or events that occurred after its release. Which is the MOST appropriate fundamental reason for this?
G23An ML team is choosing the evaluation metrics for a defect-detection (classification) model, distinguishing them from metrics for regression. Which are appropriate evaluation metrics for measuring the performance of a classification model? (Choose TWO.)
G24A company is building an AI governance program. From a responsible AI perspective, what is the term for establishing a structure that makes clear who is responsible for an AI system's decisions and why a given outcome was reached?
G24An AI team is comparing techniques to improve a chat model's responses. What is the term for the technique that uses feedback in which humans rate the quality of outputs to adjust the model's responses to align with human preferences and values?
G24A company places importance on having people oversee an AI system's behavior so that a person can intervene and stop it when undesirable behavior occurs. Which responsible AI characteristic BEST matches this idea?
G24A company is launching a generative AI chatbot for the general public and places top priority on not producing dangerous advice or harmful content. Which responsible AI characteristic BEST matches this idea?
G24A company is establishing design principles to achieve human-centered, explainable AI. Each department submitted a proposed policy. Which of the following proposals are appropriate as design principles? (Choose TWO.)
G25A company is building a generative AI application on Amazon Bedrock. Under the AWS shared responsibility model, which area is the responsibility of this company (the customer) ALONE?
G25A company wants to automatically detect whether sensitive data such as personal information is present inside the S3 buckets used for model training and inference, and to understand what needs protection. Which AWS service is the MOST suitable?
G25A company wants to improve the auditability of its AI training data. What is the concept of being able to record and track the origin and transformation history of data, that is, where it came from and what transformations it went through to reach its current form, called?
G25In a company's security training, the team is reviewing data encryption for AI systems. Which is the MOST accurate explanation of the difference between encryption at rest and encryption in transit?
G25A company is strengthening the security of an AI system that uses Bedrock. The requirements are to detect sensitive data such as personal information contained in the training data stored in S3, and to control access to AI resources with least privilege. Which TWO AWS services meet these requirements? (Choose TWO.)
G26At the start of a training session, an instructor is organizing AI-related terminology. Which of the following correctly describes the containment relationship among AI (artificial intelligence), ML (machine learning), and deep learning? Think from the broadest concept to the narrowest.
G26An ML team validated a classification model it had built and found that accuracy remained low on both the training data and the test data. The model does not seem to capture the patterns in the data well. What is this state called?
G26A power company is building a model that predicts the next day's electricity usage using temperature, day of week, and past usage records as features, to use for generation planning. Rather than judging usage with categories such as "high / low", it must output a concrete quantity that can be used directly in the plan. Which task type is MOST appropriate for this?
G26A company wants to run inference on large inputs of several hundred MB each (such as high-resolution video). Processing may take several minutes, and an immediate synchronous response is not required. The plan is to place requests in a queue and process them in order. Which inference method is MOST suitable?
G26In a training session, to organize a suitable storage destination and analysis method for each data type, a company classifies the data it handles into structured, semi-structured, and unstructured. Which of the following are unstructured data? (Choose TWO.)
G27A company is considering modernizing its NLP foundation. Across diverse natural language processing (NLP) tasks such as translation, summarization, and question answering, which characteristic MOST likely gives generative AI (a foundation model) an advantage over traditional task-specific models?
G27A company is preparing a cost estimate for using Amazon Bedrock. For on-demand use of Bedrock, what is pricing mainly based on? Which option is MOST appropriate?
G27A training session is organizing the difference between traditional machine learning, which performs prediction and classification, and generative AI. Compared with traditional machine learning, which is the MOST fundamental difference that characterizes generative AI? Choose by focusing on the nature of the output.
G27A company is estimating the development cost of a foundation model (FM). In the FM lifecycle, which is the MOST computationally expensive stage, in which the model acquires general knowledge of language and the world from a large amount of unlabeled data?
G27A company is starting to select a model to use in Amazon Bedrock. Bedrock lets you choose foundation models from multiple companies, but not every provider is available. Which of the following are appropriate examples of providers of foundation models available from Bedrock? (Choose TWO.)
G28A company performs a final safety check before releasing a generative AI chatbot. What is the name of the evaluation method in which experts take an attacker's perspective and deliberately try to elicit harmful output or vulnerabilities to surface risks before deployment?
G28An AI team is working on improving the quality of an agentic application. What is the name of the practice, increasingly emphasized in recent generative AI development, of designing the entire body of context information passed to a model (instructions, reference data, history, tool results, and so on) by deciding what to provide and in what order?
G28An AI team is organizing which challenge each Bedrock feature should be applied to. Which is the correct mapping for how to use RAG (knowledge bases), agents, guardrails, and model evaluation?
G28An AI team is looking for a way to compare several internal candidate models under fair conditions. What is the name of the practice of comparing and evaluating multiple models or methods under the same conditions using a common standard dataset and metrics?
G28An AI practitioner wants to adjust the randomness (diversity) of a generative AI's output to fit the use case. Among the inference parameters, which two directly affect the randomness of the output? (Choose TWO.)

AWS exam prep content