Zero-, one-, and few-shot learning: Practice Questions — Prompt Engineering (NVIDIA-Certified Professional: Generative AI LLMs)

Practice Questions on Zero-, One-, and Few-Shot Learning

This set of multiple-choice questions is designed to help candidates prepare for the NVIDIA-Certified Professional: Generative AI LLMs exam, focusing on the prompt engineering subtopic of zero-, one-, and few-shot learning.

1. Which of the following best describes zero-shot learning in the context of large language models?

   • A. The model is fine-tuned on a small labeled dataset before inference.
   • B. The model generates outputs based on prompts without any task-specific examples.
   • C. The model requires multiple examples in the prompt to perform the task.
   • D. The model uses reinforcement learning to improve task performance.

   Correct Answer: B

   Explanation: Zero-shot learning involves prompting the model to perform a task without providing any examples, relying solely on its pre-trained knowledge.

2. In one-shot learning, how is the prompt typically structured?

   • A. No examples are provided; only instructions are given.
   • B. Multiple examples are provided to illustrate the task.
   • C. Exactly one example is provided along with the task instructions.
   • D. The model is retrained with one example before inference.

   Correct Answer: C

   Explanation: One-shot learning uses a single example in the prompt to demonstrate the task to the model before asking it to generate the output.

3. What is a key advantage of few-shot learning over zero-shot learning?

   • A. It requires no prompt engineering.
   • B. It improves task performance by providing a few examples in the prompt.
   • C. It eliminates the need for a pre-trained model.
   • D. It always requires fine-tuning the model weights.

   Correct Answer: B

   Explanation: Few-shot learning enhances the model's ability to perform a task by including several examples in the prompt, helping it understand the task better than zero-shot prompting.

4. Which scenario best illustrates the use of few-shot learning?

   • A. Asking the model to translate text without any examples.
   • B. Providing three examples of sentiment classification before asking the model to classify a new sentence.
   • C. Fine-tuning the model on a large dataset before inference.
   • D. Using a pre-trained model without any prompt instructions.

   Correct Answer: B

   Explanation: Few-shot learning involves giving the model a small number of examples (e.g., three) in the prompt to guide its output on a similar task.

5. How does prompt engineering control model output in few-shot learning?

   • A. By changing the model architecture.
   • B. By selecting and ordering examples carefully within the prompt.
   • C. By retraining the model on the examples.
   • D. By limiting the model’s vocabulary.

   Correct Answer: B

   Explanation: In few-shot learning, prompt engineering involves crafting prompts with carefully chosen and ordered examples to influence the model’s responses effectively.

6. Why might zero-shot learning be preferred over few-shot learning in some applications?

   • A. It always yields higher accuracy.
   • B. It requires no example data and is faster to deploy.
   • C. It requires extensive fine-tuning.
   • D. It needs complex prompt engineering.

   Correct Answer: B

   Explanation: Zero-shot learning is preferred when example data is unavailable or rapid deployment is needed since it does not require providing examples in the prompt.

7. Which of the following is NOT a characteristic of one-shot learning?

   • A. Providing exactly one example in the prompt.
   • B. Model uses the example to infer task requirements.
   • C. Requires retraining the model on the example.
   • D. Helps improve performance compared to zero-shot learning.

   Correct Answer: C

   Explanation: One-shot learning does not involve retraining the model; it only provides one example in the prompt to guide the model’s output.

8. In the context of prompt engineering, what is a common challenge when using few-shot learning?

   • A. The model cannot understand multiple examples.
   • B. The prompt length limitation restricts the number of examples.
   • C. Few-shot learning eliminates the need for prompt design.
   • D. It requires model fine-tuning for each example.

   Correct Answer: B

   Explanation: One challenge in few-shot learning is the token limit of the model’s input, which restricts how many examples can be included in the prompt.

More in this topic

Zero-, one-, and few-shot learning: Quick Reference — Prompt Engineering (NVIDIA-Certified Professional: Generative AI LLMs)Prompt Engineering — NVIDIA-Certified Professional: Generative AI LLMsZero-, one-, and few-shot learning: Common Mistakes — Prompt Engineering (NVIDIA-Certified Professional: Generative AI LLMs)Controlling model output — Prompt Engineering (NVIDIA-Certified Professional: Generative AI LLMs)Zero-, one-, and few-shot learning: Worked Example — Prompt Engineering (NVIDIA-Certified Professional: Generative AI LLMs)Chain-of-thought prompting — Prompt Engineering (NVIDIA-Certified Professional: Generative AI LLMs)Domain adaptation via prompting — Prompt Engineering (NVIDIA-Certified Professional: Generative AI LLMs)Zero-, one-, and few-shot learning — Prompt Engineering (NVIDIA-Certified Professional: Generative AI LLMs)