Unverified90% confidenceFactTime unknown
RAG原始论文使用DPR（Dense Passage Retrieval）作为检索器、BART作为生成器
1
Sources
90%
Confidence
Long-term
Relevance
5/31/2026
First Seen
Sources
RAG技术入门：从大模型幻觉到检索增强生成的完整指南
bilibiliAI大模型全栈
Related Entities
Related Claims
UnverifiedRAG 框架将稠密段落检索（Dense Passage Retrieval, DPR）与生成式语言模型结合80% similarUnverified最初的RAG论文使用DPR（Dense Passage Retrieval）作为检索器，以Wikipedia作为知识源80% similarUnverifiedRAG architecture retrieves relevant document fragments from an external knowledge base before the LLM generates a response, injecting retrieved content as context into the prompt.69% similarVerifiedRAG（检索增强生成）在调用LLM前从向量数据库检索相关文档片段注入提示词，当检索文档不匹配或包含错误信息时会产生'由检索驱动的幻觉'67% similarUnverifiedRAG（检索增强生成）通过将当前输入语义向量化并在历史库中执行近似最近邻检索注入上下文窗口66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/4594API
curl https://kongchang.com/api/v1/knowledge/claims/4594MCP
get_claim(id=4594)