CCNA Oci Genai Prompt Engineering Questions

Recipe patterns are step-by-step instructions by definition. Template patterns allow reusability with placeholders for ingredients or steps. Role prompting could set persona but is not specific to recipes.

Self-consistency generates multiple reasoning paths (using a higher temperature) and then selects the most common final answer. This reduces the chance that a single flawed path leads to an incorrect result.

Output format specification explicitly tells the model the desired structure, such as JSON, XML, or markdown. The other options serve different purposes.

Using a strong system prompt with explicit constraints is the first line of defense; also, setting low temperature can reduce unpredictable outputs. But among the options, updating the system prompt with more specific guidelines is the most direct approach.

Prompt injection occurs when user input overrides the system's intended instructions. An attacker can inject 'Ignore previous instructions' to bypass safety guardrails.

Retrieval-Augmented Generation (RAG) is the most appropriate approach because it allows the chatbot to answer questions by retrieving relevant chunks from the policy documents stored in a vector store at inference time, without requiring model retraining. This decouples the knowledge base from the model weights, enabling monthly document updates by simply re-indexing the vector store, which is far more cost-effective and faster than fine-tuning or retraining.

Prompt injection occurs when user input overrides the original system instructions, potentially causing the model to ignore previous constraints and behave maliciously.

The summary task does not require example inputs; zero-shot or few-shot can work, but the most critical components are the task instruction and output format. Examples are optional and least important.

Testing with diverse inputs, A/B testing variants, and establishing evaluation criteria are key to systematic refinement.

Task instruction tells the model what to do, output format specification tells it the structure, and few-shot examples provide a concrete reference. Context/background is less critical, and temperature does not enforce schema.

Temperature and top-p (nucleus sampling) are the primary parameters that influence randomness and diversity.

The system prompt is the correct component because it is specifically designed to set the model's behavior, tone, and persona at the beginning of a conversation. In the context of a financial report generation task, a system prompt like 'You are a professional financial analyst. Use a strict, formal tone.' instructs the model to adopt that persona for all subsequent interactions, ensuring consistency across the entire session.

Cohere's document-grounded generation syntax allows you to supply search results or documents and instruct the model to answer solely from those documents, reducing hallucination.

Self-consistency runs chain-of-thought multiple times and aggregates answers (e.g., by majority vote) to improve reliability. The other options are different techniques.

Adding explicit constraints in the system prompt, setting frequency/presence penalties to reduce undesirable patterns, and testing with adversarial inputs are effective safeguards. Using high temperature or removing the system prompt would worsen the problem.

The system prompt establishes persistent behavioral guidelines that influence all subsequent interactions, ensuring consistency without repeating instructions in every user message.

RAG (Retrieval-Augmented Generation) allows the LLM to retrieve relevant document sections at inference time, so knowledge stays current without retraining. The other options either require expensive retraining for each update or lack document grounding.

Temperature scales the logits before softmax, affecting creativity. Higher values increase randomness; lower values make output more deterministic.

Few-shot prompting includes several examples of the desired input-output mapping to help the model understand the task and output format.

Providing the full document context within the prompt (e.g., using a template that includes the document text) grounds the model's response and reduces the chance it invents details. Few-shot examples can also help but are secondary to providing the source.

Explicitly specifying the output format (bullet list, max 5 items) gives the model clear constraints, reducing verbosity.

Evaluation should be based on multiple criteria beyond accuracy, especially user preference. The engineer should establish a composite metric that includes both accuracy and conciseness.

Temperature scales logits before softmax; higher values increase randomness.

A system prompt with role and constraints sets the overall behavior, and a template pattern for responses provides a consistent structure. The other options are less suitable.

Increasing temperature and top-p both increase randomness and diversity. Temperature controls the randomness of token selection, while top-p (nucleus sampling) allows a broader set of probable tokens.

RAG (Retrieval-Augmented Generation) allows the LLM to retrieve relevant document sections at inference time, so knowledge stays current without retraining. The other options either require expensive retraining for each update or lack document grounding.

Self-consistency generates several reasoning chains and aggregates the results, increasing reliability and reducing hallucinations in tasks requiring factual accuracy.

Option A is correct because the output format specification is the prompt component explicitly designed to define the structure, schema, or layout of the LLM's response. In this scenario, the prompt engineer needs the LLM to output a list of countries with specific JSON fields ('country_code' and 'name'), which is a direct instruction about the format of the output, not the task itself. This component ensures the LLM adheres to a precise data structure, such as JSON, XML, or a table, which is critical for downstream parsing or integration.

Instruction shielding (clear separation of instruction and input) and input validation/sanitization are effective defenses. Adding more examples or adjusting temperature do not address injection.

Few-shot provides examples that improve format adherence and guide the model, especially for complex tasks.

Storing templates in a version control system, using semantic versioning, and maintaining a changelog are standard practices for prompt version management. Automated testing is good but not version control per se.

Option D is correct because increasing the frequency penalty (e.g., to 0.3) directly penalizes tokens that have already appeared in the generated text, reducing the likelihood of the model repeating the same phrases. This parameter is specifically designed to discourage repetition by subtracting a fixed amount from the log-probability of each token each time it has been generated, making it the most targeted adjustment for this issue.

Chain-of-thought prompting explicitly instructs the model to show its reasoning steps, improving accuracy on multi-step problems.

Low temperature makes output more deterministic and repetitive; high temperature increases randomness and creativity.

Sanitizing prompts before submission (e.g., removing PII, using placeholders) prevents sensitive data from being sent to the model. Other options either do not prevent data leakage or are less direct.

Versioning and maintaining a library of templates are essential for tracking changes and reusability.

Option B is correct because self-consistency relies on generating multiple diverse reasoning paths via chain-of-thought (CoT) prompting. This technique samples several independent responses, each following a step-by-step reasoning process, to capture different valid approaches to the same problem.

For customer-facing applications, safety is paramount. Even if accuracy is slightly lower, ensuring no offensive content is critical to avoid reputational and legal risks. The team should prioritize safety and then work to improve accuracy.

Prompt injection vulnerabilities can cause the model to leak or repeat the prompt itself. This is a known failure mode where the model confuses the input with output.

Examples guide the model on the desired input-output pattern, improving task performance without fine-tuning.

Option B is correct because ambiguous instructions are a common prompt failure in LLM interactions. When a prompt lacks clarity or specificity, the model cannot reliably determine the user's intent, leading to off-target or inconsistent outputs. This is a fundamental failure mode in prompt engineering, as the model relies entirely on the given text to infer the desired response.

A/B testing with clear metrics (factuality, conciseness) is the standard method for comparing prompt variants. Manual inspection on a few cases is not statistically robust; other options are not comparative.

Frequency penalty reduces the likelihood of repeating tokens that have already appeared, directly combating repetition.

RAG (Retrieval-Augmented Generation) allows the LLM to retrieve relevant document sections at inference time, so knowledge stays current without retraining. The other options either require expensive retraining for each update or lack document grounding.

Frequency penalty reduces the likelihood of repeating tokens that have already appeared, directly targeting repetition. Presence penalty also helps but frequency penalty is stronger for repeated phrases.