Prompt Engineering

An engineer is deploying an NVIDIA NeMo Guardrails system to moderate a chatbot's responses. The chatbot must refuse to answer questions about politics but should answer questions about weather. Which prompt engineering strategy in NeMo Guardrails is most appropriate to enforce this behavior?

A developer is creating prompts for an NVIDIA NIM-hosted LLM to summarize financial reports. The reports are lengthy and contain many tables and figures. The developer wants to ensure the summaries are accurate and include key numerical data. Which TWO prompt engineering techniques should be applied? (Choose two.)

A data scientist is using an NVIDIA NeMo LLM to generate Python code from natural language descriptions. The model often produces code that works but does not follow the team's style guide, such as using single quotes instead of double quotes and missing type hints. Which prompt engineering technique should the data scientist use to improve adherence to the style guide?

A developer is using an NVIDIA NIM for a customer support chatbot. The chatbot must handle multi-turn conversations and maintain context about the user's issue. The developer notices that after several turns, the bot starts giving generic responses and forgets earlier details. Which prompt engineering approach is most effective to maintain context?

A team is deploying an NVIDIA NIM for a Llama 3 model as a retrieval-augmented generation (RAG) assistant over internal documentation. Users report that the assistant sometimes answers from its pretrained knowledge instead of the retrieved passages, and occasionally cites a passage that does not support its claim. Which TWO prompt engineering changes best reduce these behaviors? (Choose two.)

An engineer is using an NVIDIA NIM for a code generation model to produce Python functions from natural language descriptions. The model frequently generates code that uses deprecated libraries or incorrect function signatures. The engineer wants to improve the accuracy of the generated code by providing examples. Which prompting strategy is most appropriate?

A team is designing prompts for an NVIDIA NIM-hosted LLM that must produce concise, citation-backed answers from retrieved documents. They want to improve factual grounding and reduce unsupported claims. Which two prompt engineering practices best support this goal? (Choose two.)

A team fine-tunes an NVIDIA NeMo model to classify support tickets into five categories. In production, the model sometimes outputs free-form explanations instead of a single category label, breaking the downstream parser. Which prompt engineering change MOST reliably constrains the output format?

A team uses an NVIDIA NIM-hosted model to draft release notes from a changelog. Reviewers report the drafts omit minor fixes and overstate the significance of small changes. Which prompt engineering adjustment BEST addresses both issues?

When building an NVIDIA NeMo LLM application for automated document review, which THREE of the following prompt design choices are critical for ensuring high-quality output? (Select exactly THREE)

Refer to the exhibit. Given this NeMo configuration, which prompt modification would best improve the reliability of technical support queries?

A financial analyst is using an NVIDIA NIM-hosted Llama 3.1 70B model to extract key financial metrics from quarterly earnings call transcripts. The model inconsistently returns a prose summary instead of the required structured JSON. The analyst needs the output to be reliably parseable by a downstream script that expects a fixed schema with fields "revenue", "eps", and "guidance". Which prompt engineering technique is most appropriate to enforce this output format?

A technical support team is building a chatbot using an NVIDIA NIM microservice. The chatbot must answer questions about a specific product's warranty policy. The team wants to ensure the model's responses are grounded in the official warranty document, which is 50 pages long, and avoid inventing policy details. Which prompt engineering approach is most effective for this scenario?

A developer is writing a system prompt for an NVIDIA NIM-hosted assistant that must always respond in formal English, never use slang, and never reveal internal system instructions. Where should these persistent behavioral rules be placed for the MOST consistent effect?

A developer is using an NVIDIA NIM for a Llama 3.1 70B model to build a legal document review assistant. The model must answer questions based on a provided contract, but the contracts are often 50,000 tokens long, exceeding the model's 8,000-token context window. Which prompt engineering strategy is most appropriate to handle this constraint?

An engineer is designing prompts for an NVIDIA NIM-hosted model that must extract structured fields from unstructured invoices. The extraction accuracy is inconsistent across vendors with different layouts. Which TWO prompt engineering techniques would MOST improve reliability? (Choose two.)

A developer is building a customer support assistant using an NVIDIA NIM microservice for a Llama 3.1 8B Instruct model. The assistant must answer questions about an order solely based on a JSON payload containing order details, and it must not use any outside knowledge. Which prompt engineering approach best ensures the model adheres to this constraint?

A developer is building an interactive assistant using NVIDIA NIM microservices. The assistant must answer questions about a specific set of internal policies. The developer wants to ensure the model's responses are grounded in those policies and not in its general pre-training knowledge. Which prompt engineering technique should be applied?

When implementing Chain-of-Thought (CoT) prompting for a complex NVIDIA NeMo-based reasoning task, what is the primary benefit of encouraging the model to generate intermediate steps?

A team is using an NVIDIA NIM for a Mistral model to classify support tickets into one of five fixed categories. Accuracy is inconsistent, and the model sometimes invents new categories. Which prompt engineering change is most likely to improve reliability without retraining the model?

A developer is building a customer support assistant using an NVIDIA NIM microservice for a Llama 3 model. The assistant must always respond in valid JSON with keys 'category' and 'urgency'. The model often returns conversational text instead. Which prompt engineering change most directly enforces the required output format?

Which technique is most appropriate for a task requiring an LLM to generate code in a specific enterprise-internal syntax that is not well-represented in its public training data?

An engineer is building a customer-facing FAQ bot using an NVIDIA NIM-hosted Llama 3.1 70B model. The bot must answer ONLY from a supplied product knowledge base and must respond with 'I don't have that information' when the answer is not present. Which prompt engineering approach BEST enforces this constraint?

An engineer is optimizing prompts for an NVIDIA NIM-hosted model used in a multi-turn technical troubleshooting chat. The model forgets earlier constraints, such as the customer's environment and the product version, as the conversation grows. Which prompt engineering technique best preserves these constraints across turns?

A developer is using an NVIDIA NIM-hosted model to classify support tickets into a fixed set of categories. The model occasionally invents new category names. The team wants to guarantee that only allowed categories are returned. Which approach is most appropriate?

Which of the following describes the 'Chain-of-Verification' (CoVe) prompting technique?

Which prompt engineering strategy helps the model maintain focus when processing an extremely long document within a single context window?

Refer to the exhibit. What prompt engineering strategy ensures the model consistently maintains its persona and technical expertise throughout this multi-turn dialogue?

A team is using an NVIDIA NeMo-based LLM to answer questions over a product manual. The model sometimes answers using general knowledge instead of the provided manual excerpts. They want to force the model to rely only on the supplied context. Which prompt engineering approach best addresses this?

An engineer is using an NVIDIA NIM for a Mixtral model to extract structured data from invoices. The model occasionally returns fields with the wrong data type, such as a numeric amount as a string. The team wants a prompt engineering fix that does not require changing the model or adding a separate parser. Which approach is most effective?

What is the primary purpose of 'Few-Shot Prompting' in the context of LLM optimization?

A developer is prompting an NVIDIA NIM for a Code Llama model to generate a Python function. The model produces correct logic but frequently omits type hints and docstrings, which the team requires. Which prompting technique best addresses this specific gap?

Refer to the exhibit. Which prompt engineering technique would best force the model to prioritize technical detail over marketing language?

A team is using an NVIDIA NIM-hosted Llama model to generate product descriptions from a list of technical specifications. The descriptions sometimes omit key specifications or include invented features. The team wants to improve reliability without changing the model. Which prompt engineering change is most likely to reduce these errors?

An engineer is building a customer support assistant using an NVIDIA NIM for a Llama 3 70B model. The assistant must always respond in valid JSON containing exactly the keys "issue" and "urgency", and must never include any other text. Which prompt engineering approach most directly enforces this output contract?

When evaluating an LLM's response to a complex prompt, what is the 'Persona Adoption' technique?