A model must reason and use tools in a loop. Which technique fits?
Why C is the answer
ReAct prompting
ReAct prompting has a model reason, act with tools and observe the result in a loop.
In detail
- ReAct alternates reasoning and tool use in a loop.
- ReAct stands for reason and act.
Use case
An illustrative scenario, not a real incident.
A travel agent reasons about a trip, calls a flight tool, reads the result and reasons again.
See it in different ways
flowchart LR A["Zero-shot"] --> B["One-shot"] B --> C["Few-shot"] C --> D["Prompt chaining"] D --> E["Chain of thought and ReAct"]
mindmap
root((Prompting))
Examples
Zero-shot
One-shot
Few-shot
Persona
Role prompting
Steps
Prompt chaining
Tools
ReActstateDiagram-v2 state "Reason" as s0 state "Act" as s1 state "Observe" as s2 [*] --> s0 s0 --> s1 s1 --> s2 s2 --> s0
Key point: ReAct loops between reasoning and tools.
Why the other options are wrong
- A. Role prompting alone
- A role alone does not use tools.
- B. A lower output length
- Length has no effect on tool use.
- D. Zero-shot prompting
- Zero-shot says nothing about tools.