C_AIG_

What are some use cases for fine-tuning of a model? Note: There are 2 correct answers to this question. To sanitize model outputs. To introduce new knowledge to a model in a resource-efficient way. To quickly create iterations on a new use case. To customize outputs for specific types of inputs.

What are some examples of generative Al technologies?Note: There are 2 correct answers to this question. Al models that generate new content based on training data. Robotic process automation. Foundation models. Rule-based algorithms.

Question 3: Which of the following are grounding principles included in SAP's Al Ethics framework? Note: There are 3 correct answers to this question. Transparency and explainability. Avoid bias and discrimination. Human agency and oversight. Store all user data for legal proceedings. Maximize business profits.

What are some features of Joule? Note: There are 3 correct answers to this question. Providing coding assistance and content generation. Streamlining tasks with an Al assistant that knows your unique role. Downloading and processing data. Generating standalone applications. Maintaining data privacy while offering generative Al capabilities.

What are some functionalities provided by SAP Al Core? Note: There are 3 correct answers to this question. Orchestration of Al workflows such as model training and inference. Integration of Al services with business applications using a standardized API. Monitoring and retraining models in SAP Al Core. Continuous delivery and tenant isolation for scalability. Management of SAP S/4HANA cloud infrastructure.

Question 6: Which of the following techniques uses a prompt to generate or complete subsequent prompts (streamlining the prompt development process), and to effectively guide Al model responses?. One-shot prompting. Chain-of-thought prompting. Few-shot prompting. Meta prompting.

Question 7: What is the purpose of splitting documents into smaller overlapping chunks in a RAG system?. To reduce the storage space required for the vector database. To enable the matching of different relevant passages to user queries. To improve the efficiency of encoding queries into vector representations. To simplify the process of training the embedding model.

Question 8: What are some components of the training pipeline in SAP AI Core? Note: There are 2 correct answers to this question. The SAP HANA database for model storage. Input datasets stored in a hyperscaler object store. Automated deployment to Kubernetes clusters. Executables that define the training process.

Question 9: What does SAP recommend you do before you start training a machine learning model in SAP Al Core? Note: There are 3 correct answers to this question. Configure the model deployment in SAP AI Launchpad. Perform manual data integration with SAP HANA. Configure the training pipeline using templates. Define the required infrastructure resources for training. Register the input dataset in SAP Al Core.

Question 10: What is the primary function of the embedding model in a RAG system?. To encode queries and documents into vector representations for comparison. To generate responses based on retrieved documents and user queries. To store vector representations of documents and search for relevant passages. To evaluate the faithfulness and relevance of generated answers.

Question 11: Which technique is used to supply domain-specific knowledge to an LLM?. Prompt template expansion. Domain-adaptation training. Fine-tuning the model on general data. Retrieval-Augmented Generation.

Question 12: What are some characteristics of the SAP generative Al hub? Note: There are 2 correct answers to this question. It only supports traditional machine learning models. It provides instant access to a wide range of large language models (LLMSs). It operates independently of SAP's partners and ecosystem. It ensures relevant, reliable, and responsible business Al.

Question 13: What capabilities does the Exploration and Development feature of the generative Al hub provide? Note: There are 2 correct answers to this question. Automatic model selection. Prompt editor and management. Al playground and chat. Develop and debug ABAP code.

Question 14: What are the applications of generative Al that go beyond traditional chatbot applications? Note: There are 2 correct answers to this question. To interpret human instructions and control software systems always producing output for human consumption. To follow a specific schema - human input, Al processing, and output for human consumption. To produce outputs based on software input. To interpret human instructions and control software systems without necessarily producing output for human consumption.

Question 15: What are some benefits of SAP Business Al? Note: There are 3 correct answers to this question. Face detection and face recognition. Automatic human emotion recognition. Personalized recommendations based on Al algorithms. Intelligent business document processing. Al-powered forecasting and predictions.

Question 16: What are some SAP recommendations to evaluate pricing and rate information of model usage within SAP's generative Al hub? Note: There are 2 correct answers to this question. Use pricing models that have fixed rates irrespective of the usage patterns. Adopt best practice pricing strategies, such as outcome-based pricing. Avoid subscription-based pricing models. Weigh the cost of using advanced models against the expected return on investment.

Question 17: How can few-shot learning enhance LLM performance?. By providing a large training set to improve generalization. By offering input-output pairs that exemplify the desired behavior. By enhancing the model's computational efficiency. By reducing overfitting through regularization techniques.

Question 18: You want to download a json output for a prompt and the response. Which of the following interfaces can you use in SAP's generative Al hub in SAP AI Launchpad?. Prompt Editor. Your answer is correct Prompt management. Chat. Administration.

Question 19: How can Joule improve workforce productivity? Note: There are 2 correct answers to this question. By resolving hardware malfunctions. By offering generic task recommendations unrelated to specific roles. By maintaining strict adherence to data privacy regulations. By providing context-based role-specific task assistance.

Question 20: What are some drivers for the rapid adoption of generative AI? Note: There are 2 correct answers to this question. Significant hardware cost savings. Ease of use. Availability of skilled developers. Wide availability.

Question 21: What are the benefits of SAP's generative Al hub? Note: There are 2 correct answers to this question. Send your data to various LLM providers for training feedback. Accelerate Al development with flexible access to a broad range of models. Build custom Al solutions and extend SAP applications. Provide libraries for no-code development.

Question 22: Which of the following executables in generative Al hub works with Anthropic models?. Azure OpenAl Service. SAP Al Core. AWS Bedrock. GCP Vertex Al.

Question 23: Which of the following is unique about SAP's approach to Al?. Utilizing Al mainly for marketing purposes. Offering Al capabilities in their future products as of 2025. SAP's deep integration of Al with business processes and analytics. Focusing Al solely on customer support services.

Question 24: Which of the following is a benefit of using Retrieval Augmented Generation?. It allows LLMs to access and utilize information beyond their initial training data. It eliminates the need for fine-tuning LLMs for specific tasks. It enables LLMs to learn new languages without additional training. It reduces the computational resources required for language modeling.

Question 25: You want to assign urgency and sentiment categories to a large number of customer emails. You want to get a valid json string output for creating custom applications. You decide to develop a prompt for the same using generative Al hub. What is the main purpose of the following code in this context? prompt_test = """Your task is to extract and categorize messages. Here are some examples: {{?technique_examples}} Use the examples when extract and categorize the following message: {{?input}} Extract and return a json with the following keys and values: - "urgency" as one of {{?urgency}} - "sentiment" as one of {{?sentiment}} "categories" list of the best matching support category tags from: {{?categories}} Your complete message should be a valid json string that can be read directly and only contains the keys mentioned in t import random random.seed(42) k = 3 examples random. sample (dev_set, k) example_template = """<example> {example_input} examples '\n---\n'.join([example_template.format(example_input=example ["message"], example_output=json.dumps (example[ f_test = partial (send_request, prompt=prompt_test, technique_examples examples, **option_lists) response = f_test(input=mail["message"]). Preprocess a dataset for machine learning. Train a language model from scratch. Generate random examples for language model training. Evaluate the performance of a language model using few-shot learning.

Question 26: What advantage can you gain by leveraging different models from multiple providers through the SAP's generative Al hub?. Enhance the accuracy and relevance of Al applications that use SAP's data assets. Train new models using SAP and non-SAP data. Design new product interfaces for SAP application. Get more training data for new models.

Question 27: What is a part of LLM context optimization?. Providing the model with domain-specific knowledge needed to solve a problem. Reducing the model's size to improve efficiency. Adjusting the model's output format and style. Enhancing the computational speed of the model.

Question 28: What are some metrics to evaluate the effectiveness of a Retrieval Augmented Generation system? Note: There are 2 correct answers to this question. Relevance. Carbon footprint. Speed. Faithfulness.

Question 29: What is a Large Language Model (LLM)?. A rule-based expert system to analyze and generate grammatically correct sentences. A database system optimized for storing large volumes of textual data. A gradient boosted decision tree algorithm for predicting text. An Al model that specializes in processing, understanding, and generating human language.

Question 30: What defines SAP's approach to LLMs?. Avoiding partnerships with external AI providers. Prioritizing only the performance of open-source models. Using proprietary transformer-based models exclusively. Ensuring ethical AI practices and seamless business integration.

Question 31: What is the goal of prompt engineering?. To replace human decision-making with automated processes. To develop new neural network architectures for Al models. To optimize hardware performance for Al computations. To craft inputs that guide Al systems in generating desired outputs.

Question 32: Which neural network architecture is primarily used by LLMs?. Sequential encoder-decoder architecture. Convolutional Neural Networks (CNNs). Transformer architecture with self-attention mechanisms. Recurrent neural network architecture.