{"id":18122,"date":"2026-09-22T05:35:09","date_gmt":"2026-09-22T05:35:09","guid":{"rendered":"https:\/\/www.examlabs.com\/certification\/?p=18122"},"modified":"2026-09-22T05:35:09","modified_gmt":"2026-09-22T05:35:09","slug":"databricks-certified-generative-ai-engineer-associate-practice-test-questions-and-exam-dumps-part1-q1-20","status":"publish","type":"post","link":"https:\/\/www.examlabs.com\/certification\/databricks-certified-generative-ai-engineer-associate-practice-test-questions-and-exam-dumps-part1-q1-20\/","title":{"rendered":"Databricks Certified Generative AI Engineer Associate Practice Test Questions and Exam Dumps Part1 Q1-20"},"content":{"rendered":"

View Full <\/b>Databricks Certified Generative AI Engineer Associate Exam Dumps<\/b><\/a> and Practice Test Dumps<\/b><\/p>\n

 <\/p>\n

Question 1. Which Databricks capability is commonly used to deploy and manage generative AI model endpoints?<\/b><\/p>\n

    \n
  1. <\/b> Unity Catalog<\/span><\/li>\n
  2. <\/b> Model Serving<\/span><\/li>\n
  3. <\/b> Delta Sharing<\/span><\/li>\n
  4. <\/b> Databricks SQL Warehouse<\/span><\/li>\n<\/ol>\n

    Correct Answer: 2. Model Serving<\/b><\/p>\n

    Explanation:<\/b><\/p>\n

    Databricks Model Serving provides a managed way to deploy machine learning and generative AI models as scalable endpoints that applications can call for inference. It is useful for production workloads where applications need reliable access to deployed models without requiring teams to manually manage the underlying serving infrastructure. Generative AI applications can use serving endpoints for tasks such as text generation, summarization, classification, and other inference scenarios. Unity Catalog is primarily associated with governance and access management rather than directly serving model predictions. Therefore, Model Serving is the capability most directly associated with exposing a model through an endpoint for application inference.<\/span><\/p>\n

    Question 2. What is the primary purpose of Retrieval-Augmented Generation (RAG) in a generative AI application?<\/b><\/p>\n

      \n
    1. <\/b> To eliminate the need for an AI model<\/span><\/li>\n
    2. <\/b> To replace all structured databases<\/span><\/li>\n
    3. <\/b> To provide relevant external context to the model during generation<\/span><\/li>\n
    4. <\/b> To permanently modify the model’s training weights<\/span><\/li>\n<\/ol>\n

      Correct Answer: 3. To provide relevant external context to the model during generation<\/b><\/p>\n

      Explanation:<\/b><\/p>\n

      Retrieval-Augmented Generation combines information retrieval with generative model inference. Instead of relying exclusively on information stored within the model’s parameters, a RAG application retrieves relevant information from an external knowledge source and provides that information to the model as context. This approach is particularly useful for organization-specific information, documentation, policies, and frequently changing content. RAG does not permanently modify model weights, nor does it eliminate the need for a language model. The retrieval component supplies useful information while the generative model uses that information to formulate a response. This architecture can improve relevance and factual grounding when the retrieved information is accurate and appropriate.<\/span><\/p>\n

      Question 3. Which component converts text into numerical representations that can be used for semantic similarity search?<\/b><\/p>\n

        \n
      1. <\/b> Embedding model<\/span><\/li>\n
      2. <\/b> SQL warehouse<\/span><\/li>\n
      3. <\/b> Prompt template<\/span><\/li>\n
      4. <\/b> Web application<\/span><\/li>\n<\/ol>\n

        Correct Answer: 1. Embedding model<\/b><\/p>\n

        Explanation:<\/b><\/p>\n

        An embedding model converts text or other supported content into numerical vectors that represent semantic characteristics of the input. These vectors can be stored in a vector search system and compared with other vectors to identify semantically similar content. In a RAG workflow, documents are typically divided into chunks and transformed into embeddings during indexing. When a user submits a query, that query can also be converted into an embedding so the retrieval system can locate relevant document chunks. A prompt template does not create vector representations, while a web application generally handles the user interface and application logic. Embeddings therefore form a fundamental part of semantic retrieval.<\/span><\/p>\n

        Question 4. Why is document chunking commonly performed before creating embeddings for a RAG system?<\/b><\/p>\n

          \n
        1. <\/b> To permanently delete irrelevant documents<\/span><\/li>\n
        2. <\/b> To make every document exactly the same length<\/span><\/li>\n
        3. <\/b> To prevent users from submitting questions<\/span><\/li>\n
        4. <\/b> To create manageable, semantically meaningful retrieval units<\/span><\/li>\n<\/ol>\n

          Correct Answer: 4. To create manageable, semantically meaningful retrieval units<\/b><\/p>\n

          Explanation:<\/b><\/p>\n

          Document chunking divides large documents into smaller sections that can be embedded and retrieved independently. This is useful because a large document may contain many different topics, while a smaller chunk can represent a particular concept more precisely. During retrieval, the application can identify only the portions relevant to a user’s question and provide those portions to the language model as context. Effective chunking considers document structure, semantic boundaries, chunk size, and sometimes overlap between neighboring chunks. The purpose is not to delete information or make every document identical. Instead, chunking creates practical retrieval units that help balance context quality, retrieval precision, processing requirements, and the model’s available context window.<\/span><\/p>\n

          Question 5. What is the main role of a vector search index in a generative AI application?<\/b><\/p>\n

            \n
          1. <\/b> To execute application business logic<\/span><\/li>\n
          2. <\/b> To find records or chunks that are semantically similar to a query<\/span><\/li>\n
          3. <\/b> To generate the final natural-language response<\/span><\/li>\n
          4. <\/b> To replace model evaluation<\/span><\/li>\n<\/ol>\n

            Correct Answer: 2. To find records or chunks that are semantically similar to a query<\/b><\/p>\n

            Explanation:<\/b><\/p>\n

            A vector search index allows an application to retrieve information according to semantic similarity rather than depending only on exact keyword matches. Documents or document chunks are converted into embeddings and stored in an index. When a user submits a question, the query can be embedded and compared with stored vectors to identify potentially relevant content. The retrieved information can then be included in the prompt sent to a generative model. The vector search system itself generally does not generate the final natural-language answer. Instead, it performs the retrieval portion of the RAG architecture. Effective vector retrieval can therefore have a significant effect on the quality of the final generated response.<\/span><\/p>\n

            Question 6. Which practice can help reduce hallucinations in a RAG-based generative AI application?<\/b><\/p>\n

              \n
            1. <\/b> Providing relevant retrieved context and instructing the model to rely on it<\/span><\/li>\n
            2. <\/b> Removing all retrieved information from the prompt<\/span><\/li>\n
            3. <\/b> Increasing randomness for every response<\/span><\/li>\n
            4. <\/b> Ignoring the source documents<\/span><\/li>\n<\/ol>\n

              Correct Answer: 1. Providing relevant retrieved context and instructing the model to rely on it<\/b><\/p>\n

              Explanation:<\/b><\/p>\n

              Hallucinations occur when a generative model produces information that is unsupported, inaccurate, or invented. A properly designed RAG system can help reduce this problem by retrieving relevant information from trusted sources and providing that information to the model as context. Prompt instructions can also tell the model to use the supplied context and indicate when the available information is insufficient. However, RAG does not guarantee that hallucinations will disappear. Retrieval quality, source reliability, chunking, embedding quality, prompt design, and model behavior all affect the result. Grounding generation in relevant and trustworthy information is therefore an important technique for improving factual reliability in generative AI applications.<\/span><\/p>\n

              Question 7. What is prompt engineering primarily concerned with?<\/b><\/p>\n

                \n
              1. <\/b> Replacing all data pipelines<\/span><\/li>\n
              2. <\/b> Building physical servers<\/span><\/li>\n
              3. <\/b> Designing instructions and context that guide model behavior<\/span><\/li>\n
              4. <\/b> Creating network firewalls<\/span><\/li>\n<\/ol>\n

                Correct Answer: 3. Designing instructions and context that guide model behavior<\/b><\/p>\n

                Explanation:<\/b><\/p>\n

                Prompt engineering involves designing instructions, context, examples, constraints, and formatting that are provided to a generative AI model. A well-designed prompt can define the task, expected response format, role, relevant information, and behavioral requirements. In a RAG application, retrieved documents can also become part of the model’s context. Techniques such as few-shot examples, explicit output requirements, role instructions, and grounding directions can help make responses more consistent. Prompt engineering is different from building physical infrastructure or replacing data pipelines. Its primary purpose is to improve how the model responds to a particular task by presenting useful instructions and information in an appropriate structure.<\/span><\/p>\n

                Question 8. Which type of metric is useful for evaluating whether a retrieval system finds relevant information?<\/b><\/p>\n

                  \n
                1. <\/b> CPU utilization<\/span><\/li>\n
                2. <\/b> Recall and retrieval relevance metrics<\/span><\/li>\n
                3. <\/b> Network bandwidth<\/span><\/li>\n
                4. <\/b> Cluster startup time<\/span><\/li>\n<\/ol>\n

                  Correct Answer: 2. Recall and retrieval relevance metrics<\/b><\/p>\n

                  Explanation:<\/b><\/p>\n

                  Retrieval evaluation measures whether the search component is finding useful information for a given query. Metrics such as recall can indicate how many relevant items were successfully retrieved from the available relevant items. Other retrieval metrics can assess ranking quality, precision, and the relevance of returned results. These measurements are especially important for RAG systems because poor retrieval can result in poor answers even when the underlying language model is capable. Infrastructure measurements such as CPU utilization and cluster startup time can be useful for operational monitoring, but they do not directly measure retrieval quality. Evaluating retrieval independently helps teams identify whether an application’s weaknesses originate from search or from later generation stages.<\/span><\/p>\n

                  Question 9. What does temperature generally control in a generative language model?<\/b><\/p>\n

                    \n
                  1. <\/b> The physical temperature of the serving cluster<\/span><\/li>\n
                  2. <\/b> The size of the model’s training dataset<\/span><\/li>\n
                  3. <\/b> The number of database tables<\/span><\/li>\n
                  4. <\/b> The randomness or variability of generated outputs<\/span><\/li>\n<\/ol>\n

                    Correct Answer: 4. The randomness or variability of generated outputs<\/b><\/p>\n

                    Explanation:<\/b><\/p>\n

                    Temperature is a model-generation parameter that influences how the model selects among possible next tokens. Lower temperature settings generally encourage more deterministic outputs by favoring higher-probability choices. Higher temperature settings generally allow greater variability and can produce more diverse responses. The appropriate setting depends on the application’s requirements and should be tested against representative use cases. Temperature does not control the physical temperature of a serving cluster, the size of the training dataset, or the number of database tables. For production applications, teams may adjust temperature along with other generation parameters while evaluating response quality, consistency, factuality, and task-specific performance.<\/span><\/p>\n

                    Question 10. Which Databricks capability provides centralized governance for data and AI assets?<\/b><\/p>\n

                      \n
                    1. <\/b> Unity Catalog<\/span><\/li>\n
                    2. <\/b> Notebook comments<\/span><\/li>\n
                    3. <\/b> Model temperature<\/span><\/li>\n
                    4. <\/b> Prompt templates<\/span><\/li>\n<\/ol>\n

                      Correct Answer: 1. Unity Catalog<\/b><\/p>\n

                      Explanation:<\/b><\/p>\n

                      Unity Catalog provides centralized governance capabilities for data and AI assets within Databricks environments. It can help organizations manage access permissions, organize resources, support discovery, and provide governance and lineage capabilities across supported assets. Governance becomes especially important for generative AI applications that may access enterprise documents, models, vector indexes, and sensitive information. Prompt templates and model parameters primarily influence application behavior rather than centralized asset governance. By using a centralized governance framework, organizations can apply consistent access controls and auditing practices across workloads. This supports more controlled development and operation of generative AI applications while helping teams maintain appropriate oversight of data and AI resources.<\/span><\/p>\n

                      Question 11. What is the purpose of a centralized AI access layer or gateway?<\/b><\/p>\n

                        \n
                      1. <\/b> To physically manufacture GPUs<\/span><\/li>\n
                      2. <\/b> To centralize controls such as access, monitoring, and model usage<\/span><\/li>\n
                      3. <\/b> To remove all application prompts<\/span><\/li>\n
                      4. <\/b> To convert every model into a database<\/span><\/li>\n<\/ol>\n

                        Correct Answer: 2. To centralize controls such as access, monitoring, and model usage<\/b><\/p>\n

                        Explanation:<\/b><\/p>\n

                        A centralized AI access layer can provide a consistent interface between applications and the models they use. Depending on the platform configuration, this layer may support access control, usage monitoring, rate management, logging, and other governance or operational controls. Centralization can be particularly useful when an organization operates multiple AI applications or uses several model providers. Instead of implementing identical controls independently within every application, common policies can be managed through a shared access layer. This does not eliminate prompts or convert models into databases. Its purpose is to provide controlled, observable, and manageable access to AI capabilities across different applications and workloads.<\/span><\/p>\n

                        Question 12. In a RAG pipeline, what should generally happen before the application generates an answer?<\/b><\/p>\n

                          \n
                        1. <\/b> The model should permanently retrain itself<\/span><\/li>\n
                        2. <\/b> The source data should always be deleted<\/span><\/li>\n
                        3. <\/b> Relevant information should be retrieved and supplied as context<\/span><\/li>\n
                        4. <\/b> All model parameters should be manually edited<\/span><\/li>\n<\/ol>\n

                          Correct Answer: 3. Relevant information should be retrieved and supplied as context<\/b><\/p>\n

                          Explanation:<\/b><\/p>\n

                          A typical RAG workflow begins with a user query and uses a retrieval mechanism to identify information relevant to that query. The retrieved content is then incorporated into the model’s input, usually through a prompt or context structure. The generative model uses the supplied information together with its learned capabilities and instructions to produce an answer. This approach allows an application to use organization-specific or newly updated information without necessarily retraining the model whenever source documents change. RAG does not require manually editing model parameters for each new document. Instead, the knowledge source and retrieval index can be updated so that new information becomes available during inference.<\/span><\/p>\n

                          Question 13. What is the purpose of evaluating a generative AI application before production deployment?<\/b><\/p>\n

                            \n
                          1. <\/b> To verify quality, safety, relevance, and expected behavior<\/span><\/li>\n
                          2. <\/b> To guarantee that the model will never make mistakes<\/span><\/li>\n
                          3. <\/b> To eliminate the need for monitoring<\/span><\/li>\n
                          4. <\/b> To increase model size automatically<\/span><\/li>\n<\/ol>\n

                            Correct Answer: 1. To verify quality, safety, relevance, and expected behavior<\/b><\/p>\n

                            Explanation:<\/b><\/p>\n

                            Evaluation allows teams to determine whether a generative AI application meets predefined requirements before it is released to users. Depending on the application, evaluation may examine response relevance, factual grounding, retrieval quality, safety, correctness, latency, and adherence to required output formats. Automated tests can be combined with human evaluation when qualitative judgment is needed. Evaluation cannot guarantee that a model will never make mistakes, and it does not remove the need for production monitoring. Instead, evaluation provides evidence about application behavior under controlled conditions and helps identify weaknesses that should be addressed before deployment. A strong evaluation set should contain representative examples of the application’s expected real-world workloads.<\/span><\/p>\n

                            Question 14. Why are guardrails useful in generative AI applications?<\/b><\/p>\n

                              \n
                            1. <\/b> They increase storage capacity<\/span><\/li>\n
                            2. <\/b> They replace vector indexes<\/span><\/li>\n
                            3. <\/b> They prevent all model updates<\/span><\/li>\n
                            4. <\/b> They help constrain unsafe, invalid, or unwanted behavior<\/span><\/li>\n<\/ol>\n

                              Correct Answer: 4. They help constrain unsafe, invalid, or unwanted behavior<\/b><\/p>\n

                              Explanation:<\/b><\/p>\n

                              Guardrails are controls that help reduce unwanted behavior from a generative AI application. They may include input validation, output filtering, content safety controls, topic restrictions, structured output requirements, and application-specific policies. For example, an enterprise assistant might be configured to stay within approved topics or avoid exposing sensitive information. Guardrails do not guarantee perfect safety because AI systems can encounter unexpected inputs and complex situations. They should therefore be combined with evaluation, monitoring, access controls, secure data handling, and appropriate human oversight. Their role is to place additional constraints around model behavior so the application is more likely to operate within its intended boundaries.<\/span><\/p>\n

                              Question 15. What is a benefit of storing source documents and retrieved content with appropriate metadata in a RAG system?<\/b><\/p>\n

                                \n
                              1. <\/b> It guarantees every generated answer is correct<\/span><\/li>\n
                              2. <\/b> It supports filtering, traceability, and improved retrieval<\/span><\/li>\n
                              3. <\/b> It eliminates the need for embeddings<\/span><\/li>\n
                              4. <\/b> It prevents all document updates<\/span><\/li>\n<\/ol>\n

                                Correct Answer: 2. It supports filtering, traceability, and improved retrieval<\/b><\/p>\n

                                Explanation:<\/b><\/p>\n

                                Metadata provides additional information about documents and document chunks in a RAG system. Examples can include document identifiers, source locations, timestamps, categories, ownership information, and access-related attributes. Metadata can support filtering so that retrieval considers only appropriate content. It can also help applications identify the source of retrieved information, which improves traceability and makes troubleshooting easier. Metadata does not guarantee that every generated answer will be correct, and it does not eliminate the need for embeddings. Instead, metadata complements semantic retrieval by providing structured information that can improve filtering, organization, governance, and understanding of where retrieved context originated.<\/span><\/p>\n

                                Question 16. Which approach is generally appropriate when an application needs structured, machine-readable model output?<\/b><\/p>\n

                                  \n
                                1. <\/b> Requesting a specific output schema or structured format<\/span><\/li>\n
                                2. <\/b> Removing all instructions from the prompt<\/span><\/li>\n
                                3. <\/b> Increasing temperature as much as possible<\/span><\/li>\n
                                4. <\/b> Providing unrelated documents as context<\/span><\/li>\n<\/ol>\n

                                  Correct Answer: 1. Requesting a specific output schema or structured format<\/b><\/p>\n

                                  Explanation:<\/b><\/p>\n

                                  Generative AI applications frequently need model responses that downstream software can process reliably. When structured output is required, developers can define a specific schema and clearly communicate the expected fields and format to the model. Depending on the model and platform capabilities, structured-output features can provide additional enforcement of the required format. Removing instructions or adding unrelated context generally makes consistent structured generation more difficult. Increasing temperature is also not inherently beneficial because greater variability can introduce formatting inconsistencies. Clear schemas, appropriate model capabilities, and validation of generated responses are useful techniques for integrating generative AI with deterministic application workflows that depend on machine-readable output.<\/span><\/p>\n

                                  Question 17. What does grounding mean in the context of generative AI?<\/b><\/p>\n

                                    \n
                                  1. <\/b> Increasing the model’s parameter count<\/span><\/li>\n
                                  2. <\/b> Connecting generated responses to reliable external information or evidence<\/span><\/li>\n
                                  3. <\/b> Running the model only on physical machines<\/span><\/li>\n
                                  4. <\/b> Removing all context from the prompt<\/span><\/li>\n<\/ol>\n

                                    Correct Answer: 2. Connecting generated responses to reliable external information or evidence<\/b><\/p>\n

                                    Explanation:<\/b><\/p>\n

                                    Grounding refers to connecting a model’s generated response to relevant information or evidence supplied from an external source. In enterprise RAG applications, grounding commonly involves retrieving approved documents or other knowledge sources and including the relevant information in the model’s context. This can help the model answer questions using organization-specific information rather than relying exclusively on knowledge encoded in its parameters. Grounding can also support traceability when source references are retained or displayed. It does not mean increasing model size or removing context. Grounding can improve factual reliability, but its effectiveness still depends on the quality of the source information, retrieval process, prompt design, and evaluation strategy.<\/span><\/p>\n

                                    Question 18. What is an important consideration when selecting a foundation model for a generative AI application?<\/b><\/p>\n

                                      \n
                                    1. <\/b> Only the model’s name<\/span><\/li>\n
                                    2. <\/b> Whether the model has the largest possible parameter count<\/span><\/li>\n
                                    3. <\/b> Task requirements, quality, latency, cost, and supported capabilities<\/span><\/li>\n
                                    4. <\/b> Whether the model requires no evaluation<\/span><\/li>\n<\/ol>\n

                                      Correct Answer: 3. Task requirements, quality, latency, cost, and supported capabilities<\/b><\/p>\n

                                      Explanation:<\/b><\/p>\n

                                      Foundation model selection should be based on the actual requirements of the intended application. Teams may consider model quality for relevant tasks, context-window requirements, language support, instruction-following capability, latency, throughput, inference cost, licensing considerations, and integration requirements. Security and governance requirements can also influence the decision. A model with the largest parameter count is not automatically appropriate because another model may provide sufficient quality with lower latency or cost. Every selected model should also be evaluated against representative application workloads. A systematic evaluation process helps teams understand whether the model provides the necessary quality and operational characteristics for the particular generative AI application.<\/span><\/p>\n

                                      Question 19. Why is monitoring important after deploying a generative AI application?<\/b><\/p>\n

                                        \n
                                      1. <\/b> Production behavior can change, and issues may not appear during testing<\/span><\/li>\n
                                      2. <\/b> Monitoring permanently improves the model’s training data<\/span><\/li>\n
                                      3. <\/b> Monitoring eliminates the need for security controls<\/span><\/li>\n
                                      4. <\/b> Production applications never require maintenance<\/span><\/li>\n<\/ol>\n

                                        Correct Answer: 1. Production behavior can change, and issues may not appear during testing<\/b><\/p>\n

                                        Explanation:<\/b><\/p>\n

                                        Production monitoring helps teams understand how a generative AI application behaves under real-world usage. Important signals can include latency, request volume, errors, resource utilization, retrieval performance, response quality, and safety-related events. Some problems may not appear during controlled testing because production inputs and usage patterns can differ significantly from evaluation datasets. Monitoring therefore helps teams identify operational failures, unexpected behavior, regressions, or changes in application performance. Monitoring does not automatically improve training data and cannot replace security controls. A production generative AI application should be continuously observed so that teams can investigate issues, evaluate changes, and make controlled improvements when necessary.<\/span><\/p>\n

                                        Question 20. Which sequence best represents a practical production workflow for a RAG-based generative AI application?<\/b><\/p>\n

                                          \n
                                        1. <\/b> Generate an answer first, then create the knowledge base afterward<\/span><\/li>\n
                                        2. <\/b> Deploy immediately and evaluate only if users complain<\/span><\/li>\n
                                        3. <\/b> Delete source documents after creating embeddings<\/span><\/li>\n
                                        4. <\/b> Prepare and index data, retrieve relevant context, generate responses, evaluate, monitor, and improve<\/span><\/li>\n<\/ol>\n

                                          Correct Answer: 4. Prepare and index data, retrieve relevant context, generate responses, evaluate, monitor, and improve<\/b><\/p>\n

                                          Explanation:<\/b><\/p>\n

                                          A practical RAG production workflow contains several connected stages. Source documents are prepared, cleaned, chunked, enriched with appropriate metadata, and converted into embeddings for retrieval. When a user submits a query, the system retrieves relevant information and incorporates that context into the model input. The generated response can then be evaluated for relevance, grounding, quality, safety, and operational characteristics. After deployment, monitoring helps identify unexpected behavior, performance changes, and retrieval or generation problems. Feedback from evaluation and monitoring can be used to improve prompts, retrieval strategies, source data, model selection, or other application components. This lifecycle supports a controlled, maintainable, and continuously improving generative AI application.<\/span><\/p>\n

                                           <\/p>\n","protected":false},"excerpt":{"rendered":"

                                          View Full Databricks Certified Generative AI Engineer Associate Exam Dumps and Practice Test Dumps   Question 1. Which Databricks capability is commonly used to deploy and manage generative AI model endpoints? Unity Catalog Model Serving Delta Sharing Databricks SQL Warehouse Correct Answer: 2. Model Serving Explanation: Databricks Model Serving provides a managed way to deploy […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[1648,1647],"tags":[],"_links":{"self":[{"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/posts\/18122"}],"collection":[{"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/comments?post=18122"}],"version-history":[{"count":1,"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/posts\/18122\/revisions"}],"predecessor-version":[{"id":18123,"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/posts\/18122\/revisions\/18123"}],"wp:attachment":[{"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/media?parent=18122"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/categories?post=18122"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.examlabs.com\/certification\/wp-json\/wp\/v2\/tags?post=18122"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}