When designing a prompt for a high-stakes financial application, which TWO techniques are most effective for mitigating the risk of prompt injection and ensuring the model adheres to its system-defined guardrails?

Using distinct delimiters for user input

Implementing recursive character splitting

Placing user input before the system instructions

Utilizing a robust system-level prompt

A data scientist is using few-shot prompting to classify customer feedback into five distinct categories. Which THREE factors are most critical for selecting effective examples to include in the prompt?

Semantic similarity to the target input

Alphabetical ordering of examples

Total character count of all examples

A developer wants to implement a 'Persona' or 'Role-playing' prompt to improve a customer service bot. Which TWO benefits does this technique provide for the model's output?

Consistent tone and stylistic alignment

Reduction in total inference latency

Implicit context for domain-specific jargon

Elimination of the need for stop sequences

Automatic verification of factual accuracy

When utilizing the ReAct (Reason + Act) prompting framework, what is the primary purpose of the 'Observation' step in the model's iterative loop?

It generates the final answer for the user

It provides the model with the result of an action

It defines the persona for the reasoning trace

It is used to penalize repetitive reasoning paths

An engineer is struggling with 'hallucinations' in a model that summarizes legal documents. Which TWO prompt engineering adjustments are most likely to reduce the occurrence of fabricated information?

Instructing the model to say 'I don't know' if unsure

Adding a requirement for direct quotes and citations

Increasing the Top-k parameter to 100

Removing all few-shot examples from the prompt

Using a very high temperature setting

NCP-GENL Prompt Engineering — Timed 10 Questions

Time remaining

10:00

Exam-pace drill

Allow 1 minute per question. On the real NCP-GENL exam you have approximately 72 seconds per question — this session trains you to maintain that pace under pressure.

NCP-GENLTimed Practice

1 / 1010%

10:00

Prompt Engineeringmedium

A developer is building a technical support bot that must analyze complex log files to identify root causes. The initial zero-shot prompt often yields generic summaries that miss specific error codes. Which strategy should the developer implement to ensure the model performs logical deduction step-by-step before providing a final answer?

10 minute time limit — choose an answer to begin.

10 questions · 10 minute exam-pace drill.

Time remaining

10:00

Exam-pace drill

Allow 1 minute per question. On the real NCP-GENL exam you have approximately 72 seconds per question — this session trains you to maintain that pace under pressure.

NCP-GENLTimed Practice

1 / 1010%

10:00

Prompt Engineeringmedium

10 minute time limit — choose an answer to begin.

10 questions · 10 minute exam-pace drill.