Exams / AIF-C01 / Choosing FMs and inference parameters / Question 3
AIF-C01 practice question: Choosing FMs and inference parameters (3 of 4)
Applications of Foundation Models A free question from our original AIF-C01 practice exam. Try it, then open the answer to see why each option is right or wrong.
A tax advisory firm's assistant sends the same 40-page tax guide, followed by a short customer question, to a model on Amazon Bedrock thousands of times a day. Responses are slow and input costs are high. The guide changes only once a year. Which feature addresses BOTH problems with the least change?
Choose one answer.
- AA. Prompt caching, with the static guide placed at the start of the prompt
- BB. A higher temperature, so the model reads the guide more quickly
- CC. Continued pre-training on the guide every time a question arrives
- DD. A larger maximum response length, so fewer requests are needed
Show the answer and explanation
Answer: A
- A. Right: Correct. Caching a repeated, unchanged prompt prefix means it does not have to be reprocessed, which cuts latency and input token cost.
- B. Wrong: Temperature affects randomness, not processing speed or token cost.
- C. Wrong: Training on every request would be slow and very expensive. Customisation is not a per-request step.
- D. Wrong: Longer responses add output tokens and cost. They do not reduce the cost of resending the guide.
Explanation
Prompt caching on Amazon Bedrock stores a static prompt prefix, such as a long document, system prompt or tool definitions, so later requests that start with the same content can reuse it. Cached reads are billed at a lower rate and need less processing. Static content should come first and changing content last, because any change to the prefix causes a cache miss.
Study
Keep practising
Answer questions with a score at the end, or practise just this topic with the explanation after each one. No sign-up needed.