A Generative AI Engineer has been reviewing issues with their company's LLM based question- answering assistant and has determined that a technique called prompt chaining could help alleviate some…

Question

Options

• AIt allows you to break down complex tasks into multiple independent subtasks. This enables the
• BIt allows you to reduce the latency of your applications. By having multiple chains participating in
• CIt allows you to decrease the effort involved in crafting a prompt. Chains make it possible to reuse
• DIt reduces the average cost of a typical request. Chains make more efficient use of the tokens

• A
  93% (43)
• B
  2% (1)
• C
  4% (2)

Explanation

Prompt chaining is a technique where a complex task is decomposed into a sequence of smaller, more focused subtasks. Each subtask is handled by a dedicated prompt, and the output of one prompt becomes the input (context) for the next. For a Q&A assistant, this might mean one prompt identifies the user intent, a second retrieves relevant context, and a third generates the final answer. This improves accuracy and reasoning quality because each prompt in the chain has a narrower, well-defined responsibility. Option B is incorrect because chaining typically adds latency, not reduces it. Option C is incorrect because chaining increases prompt complexity, not decreases it. Option D is incorrect because chaining generally uses more tokens overall, not fewer.

No community discussion yet for this question.