Expired65% confidenceFactExact time
RAG (Retrieval-Augmented Generation) was proposed by Meta in 2020
3
Sources
65%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
Valid until: 9/30/2026(expired)
Sources
Agent Harness: A New AI Agent Paradigm from Prompt Engineering to Execution Environment Orchestration
bilibili认真的笨笨6/21/2026
Related Claims
UnverifiedRAG（Retrieval-Augmented Generation）是2023年以来最主流的知识增强方案78% similarUnverifiedRAG由Meta AI的Lewis等人于2020年正式提出，最初架构用密集向量检索器DPR从Wikipedia召回文档段落再喂给seq2seq生成器71% similarPartially VerifiedRAG (Retrieval-Augmented Generation) was the dominant paradigm for LLM applications in 2023, solving issues of outdated model knowledge and hallucinations by injecting retrieval results from external knowledge bases into the model's context.69% similarUnverifiedThe core idea of Modular RAG comes from a review paper of the same name published in 2023, which abstracts the RAG system into three major modules: indexing, retrieval, and generation.69% similarUnverifiedRAG由Meta研究团队于2020年首次提出，核心思想是将参数化记忆与非参数化记忆结合68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/42710API
curl https://kongchang.com/api/v1/knowledge/claims/42710MCP
get_claim(id=42710)