NCP-GENL Prompt Engineering Practice Question
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.)
⚠ Common exam trap
The trap here is thinking that chain-of-thought or high temperature will improve numerical accuracy, when the real need is explicit instructions and examples that enforce exact inclusion of figures.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Instruct the model to extract and include all monetary values and percentages exactly as they appear in the report, and to avoid rounding or paraphrasing numbers.
To ensure accurate financial summaries with key numerical data, the developer should explicitly instruct the model to include all monetary values and percentages exactly as they appear, and provide a few-shot example demonstrating correct inclusion of numbers. These two techniques directly guide the model to preserve numerical accuracy and follow the desired format. High temperature, chain-of-thought, or ignoring tables would not achieve the goal.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Use a chain-of-thought prompt that asks the model to first list all tables and figures, then write the summary.
Why it's wrong here
Chain-of-thought can help with reasoning, but for summarization of lengthy reports with tables, it may not directly improve numerical accuracy. Listing tables and figures first could consume context and may not ensure that the final summary includes the correct numbers. Moreover, the model might still omit or misstate figures in the summary. The more direct techniques are explicit instructions and examples.
- ✗
Set the temperature to a high value like 1.0 to encourage the model to creatively interpret the financial data and provide insightful analysis.
Why it's wrong here
High temperature increases randomness and creativity, which is detrimental for financial summarization where factual accuracy is critical. The model might invent or misstate numbers, or provide speculative analysis not grounded in the report. For accurate summaries that include exact numerical data, a low temperature (e.g., 0.1 or 0.2) is more appropriate to reduce variability and hallucination.
- ✓
Instruct the model to extract and include all monetary values and percentages exactly as they appear in the report, and to avoid rounding or paraphrasing numbers.
Why this is correct
This instruction directly addresses the need for accurate numerical data. By explicitly telling the model to include all monetary values and percentages exactly as they appear, and to avoid rounding or paraphrasing, you reduce the risk of the model altering numbers. It sets a clear constraint that the model must adhere to, which is crucial for financial summaries where precision is paramount.
- ✗
Instruct the model to ignore any tables and figures and focus only on the narrative text to avoid confusion.
Why it's wrong here
Ignoring tables and figures would omit key numerical data, which contradicts the goal of including key numerical data in the summary. Financial reports often convey critical information through tables, so excluding them would lead to incomplete and potentially misleading summaries. This instruction is counterproductive to the requirement of accuracy and inclusion of numerical data.
- ✓
Provide a few-shot example of a summary that correctly includes key numbers from a sample report, demonstrating the desired format and level of detail.
Why this is correct
A few-shot example shows the model exactly what a good summary looks like, including how to incorporate numerical data. By demonstrating the format and detail, the model can mimic the pattern for new reports. This is especially useful for complex documents with tables, as the example can illustrate how to extract and present figures accurately, reinforcing the instruction to include all key numbers.
About these practice questions
Courseiva writes every NCP-GENL question from scratch — 352 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official NVIDIA exam blueprint
This NCP-GENL practice question is part of Courseiva's free NVIDIA certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the NCP-GENL exam.