What does ReAct stand for?

ReAct stands for 'Reasoning and Acting,' referring to its core technique of interleaving explicit reasoning steps with concrete tool or environment actions in a single prompting loop.

Who introduced ReAct?

ReAct was introduced in the 2022 paper 'ReAct: Synergizing Reasoning and Acting in Language Models' by Shunyu Yao and colleagues, evaluating it on tasks including question answering, fact verification, and interactive text-based environments.

What is the Thought-Action-Observation loop?

It's the core repeating cycle in ReAct: the model produces a 'Thought' (reasoning about what to do next), takes an 'Action' (such as a tool call), receives an 'Observation' (the result of that action), and then generates a new Thought based on the updated information, repeating until a final answer is ready.

How does ReAct reduce hallucination?

By grounding the model's reasoning in results retrieved from real actions (like search queries or database lookups) rather than relying solely on the model's internal, potentially outdated or incorrect knowledge, ReAct gives the model a way to verify and update its beliefs mid-reasoning.

Is ReAct still used in modern agent frameworks?

Yes, the Thought-Action-Observation pattern remains foundational to how many agentic systems and frameworks are designed, even when not explicitly labeled 'ReAct,' since it's a natural template for any system that must reason about when and how to use tools.

How is ReAct different from plain function calling?

Function calling is the underlying mechanism that lets a model invoke external functions. ReAct is a higher-level prompting pattern that structures how and when those calls happen within an explicit, iterative reasoning loop, rather than just describing the calling mechanism itself.

AI Fundamentals

ReAct Prompting

IntermediateTechnique6.5K learners

ReAct (Reasoning and Acting) is a prompting framework that interleaves a language model's step-by-step reasoning with concrete actions, such as tool or API calls, letting the model reason about what information it needs, take an action to…

#ReActPrompting #AIFundamentals#Technique#Intermediate#ChainOfThoughtPrompting #ToolUse #FunctionCalling #MultiAgentSystem #ArtificialIntelligence#Glossary#SkillVeris

Definition

Overview

Prior to ReAct, prompting research had largely explored reasoning (chain-of-thought, generating intermediate reasoning steps toward an answer) and acting (function/tool calling, letting a model invoke external tools) as somewhat separate techniques. ReAct, introduced in a 2022 paper by Shunyu Yao and colleagues, proposed combining them into a single interleaved loop: the model generates a reasoning trace ('Thought: I need to find the current population of X'), takes an action based on that reasoning ('Action: search("current population of X")'), receives an observation from executing that action (the tool's returned result), and then continues reasoning based on the new information, repeating this Thought-Action-Observation cycle until it has gathered enough information to produce a final answer. The key insight is that reasoning and acting are mutually reinforcing: reasoning helps the model decide what action to take and how to interpret an action's result, while taking actions grounds the model's reasoning in real, verifiable information rather than relying entirely on its own possibly outdated or hallucinated internal knowledge. The paper demonstrated that this combination reduced hallucination and error propagation compared to reasoning-only approaches (which cannot verify facts against the real world) and improved interpretability compared to acting-only approaches (where it's unclear why the model chose a given action), on tasks spanning question answering, fact verification, and interactive decision-making in simulated environments. ReAct has become one of the foundational prompting patterns underlying modern LLM agents: essentially any agentic system that calls tools iteratively — a coding agent that reasons about what file to inspect next, a research agent that searches, reads, and refines its query, or a customer-support agent that looks up account details before answering — implements some variation of the ReAct Thought-Action-Observation loop, whether or not it's explicitly labeled as such. It remains a standard reference pattern taught alongside function calling and tool use when designing agentic systems, and many agent frameworks (LangChain's agent executors, for instance) originally implemented ReAct as a default reasoning strategy.

Key Concepts

Interleaves reasoning ('Thought') with concrete actions ('Action') and their results ('Observation')
Introduced in a 2022 paper by Shunyu Yao and colleagues
Grounds model reasoning in real, verifiable information via tool/API calls
Reduces hallucination compared to reasoning-only approaches with no external verification
Improves interpretability compared to acting-only approaches with no visible reasoning trace
Forms the basis of the Thought-Action-Observation loop used in many agent frameworks
Applicable to question answering, fact verification, and interactive decision-making tasks
A foundational pattern underlying most modern iterative, tool-using LLM agents

Use Cases

Building research agents that iteratively search, read, and refine queries to answer questions

Coding agents that reason about which file or command to inspect next before acting

Fact-checking or verification pipelines that ground claims in retrieved evidence

Customer-support agents that look up account or order details before responding

Interactive decision-making agents operating in simulated or real environments

General-purpose agent frameworks implementing an iterative reasoning-action loop

Frequently Asked Questions

From the Blog

Projects & Case Studies

Build a To-Do List App in React

A comprehensive guide to build a to-do list app in react — written for learners at every level.

Project: Build a Full-Stack To-Do App with React, Node.js and MongoDB

A full-stack to-do app is the perfect first MERN project — it covers every concept you'll use in production: REST APIs, database CRUD operations, JWT authentication, and deploying a frontend and backend separately. Build it once, understand the full stack.

React Hooks Explained: useState, useEffect, and Beyond

React Hooks replaced class components and changed how React developers think about state and side effects. This guide explains useState, useEffect, useContext, useRef, and custom hooks clearly, with practical examples for each.

Learn React Through Game Development: Build a Chess Clock

A chess clock is the perfect React project — it's small enough to finish in an afternoon but teaches useState, useEffect, timer management, and conditional rendering in a context that makes every concept feel purposeful. No boring counter apps.