Unverified50% confidenceFactExact time
检索增强生成（RAG）范式由Lewis等人于2020年在Meta AI提出，核心思想是将参数化知识与非参数化知识相结合
1
Sources
50%
Confidence
Long-term
Relevance
7/14/2026
First Seen
Sources
Adaptive Recall：基于MCP协议的AI助手持久记忆方案解析
Related Claims
UnverifiedRAG检索增强生成由Lewis等人于2020年在Facebook AI Research提出，核心在于将参数化知识与非参数化知识解耦，使知识更新无需重新训练模型78% similarUnverifiedThe industry best practice is to combine fine-tuning with RAG—first fine-tuning to give the model domain understanding capability, then supplementing the latest dynamic knowledge through RAG.77% similarUnverifiedCompared to baking knowledge into model parameters, RAG supports real-time knowledge base updates and is the mainstream solution for enterprise-level AI applications to address the problem of knowledge staleness.76% similarUnverifiedRAG (Retrieval-Augmented Generation) was proposed by Lewis et al. in 2020, with the core idea of decoupling parametric knowledge from non-parametric knowledge and dynamically retrieving relevant documents to concatenate into the prompt at inference time.76% similarUnverified传统大模型依赖参数化知识（将文本编码进模型权重参数），而RAG引入非参数化知识（将外部文档库作为可实时检索的动态知识源）76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/503720API
curl https://kongchang.com/api/v1/knowledge/claims/503720MCP
get_claim(id=503720)