Verified65% confidenceFactExact time
The ReAct paradigm proposed by Yao et al. in 2023 has the model alternately perform Thought and Action steps, establishing the basic working mode of modern Agents.
3
Sources
65%
Confidence
Long-term
Relevance
8/26/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedReAct (Reasoning + Acting) is an Agent paradigm proposed by Google in 2022, whose core idea is to have large models alternate iteratively between reasoning and acting.79% similarVerifiedThe ReAct (Reasoning + Acting) paper published in 2022 demonstrated that having LLMs alternate between reasoning and acting can significantly improve task completion rates, laying the design foundation for current mainstream agent frameworks.77% similarUnverified现代Agent框架如LangChain、AutoGPT、Claude的Computer Use通过ReAct（Reasoning + Acting）范式让模型观察环境、推理并采取行动形成闭环76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/803028API
curl https://kongchang.com/api/v1/knowledge/claims/803028MCP
get_claim(id=803028)