Prepare and test your skills
Prepare and test your skills
Free sample question, answered and explained below. The Generative AI Leader practice bank has 2,191 more like it. Try 10 free: instant score, no account.
A financial analyst team is using a large language model on Google Cloud to solve complex multi-step math and deductive logic problems from corporate earnings reports. When prompted directly for a final numerical calculation, the model frequently makes logical errors and arrives at incorrect conclusions.
Which prompting technique should the team implement to improve accuracy by having the model decompose the problem into intermediate logical steps before answering?
Chain-of-thought (CoT) prompting is an advanced prompt engineering technique that guides a large language model to articulate intermediate reasoning steps before arriving at a final answer. By breaking down a complex problem into a sequential series of logical statements, the model allocates compute to step-by-step reasoning rather than jumping immediately to a conclusion.
Compared to asking for a direct answer or using simple formatting prefixes, chain-of-thought prompting directly tackles the underlying limitation of autoregressive language models when solving multi-layered logical and arithmetic tasks.
Negative prompt constraint filtering involves specifying tokens, topics, or formatting elements that the model must exclude from its generation (for example, instructing an image or video model to avoid rendering specific artifacts or text).
Prefix-only formatting is a structural prompt design method where labels followed by colons (such as TASK:, CONTEXT:, or OBJECTS:) are used to organize distinct sections of a prompt and guide simple classification or extraction tasks.
Large vs. Small) using explicit header labels in the prompt structure.Zero-shot direct answer prompting presents a task to the language model without demonstration examples and requests an immediate final answer without requiring intermediate reasoning or explanation.
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated