CH 02 · ReAct, Reasoning and Acting Together · 01/06
Read as one pageReAct, Reasoning and Acting Together
ReAct interleaves thought, action, and observation in a loop so a model can call tools for real information, curbing hallucination in modern AI agents.
6 slides · 3 min read
In this chapter
Paper: ReAct: Synergizing Reasoning and Acting in Language Models (2022)
Chapter 1 gave the model an inner monologue. But thinking alone has a limit: a model can only reason about what it already knows. Ask it "What is the weather in Tokyo right now?" and no amount of step-by-step thinking will help, because the answer is not inside it. ReAct closes this gap by letting the model act: look things up, use tools, and bring real information back into its reasoning. This paper is the blueprint for nearly every AI agent built since.
Finished this idea?
Your learning progress stays in this browser.