Unverified50% confidenceFactExact time
Representative models for dense vector retrieval include DPR, BGE, and text-embedding-ada.
1
Sources
50%
Confidence
Long-term
Relevance
9/11/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedBGE和text-embedding-3是当前可选的向量嵌入（Embedding）模型71% similarUnverifiedVector databases use embedding models (such as OpenAI's text-embedding-3-large or the open-source BGE-M3) to convert text into floating-point vectors of hundreds to thousands of dimensions, where vector distance corresponds to semantic similarity.68% similarUnverifiedRAG 框架将稠密段落检索（Dense Passage Retrieval, DPR）与生成式语言模型结合68% similarUnverified对比学习在文本检索领域的突破来自2021年的DPR和SimCSE；DPR验证了双塔结构+批内负例策略在开放域问答检索上可超越BM2568% similarUnverified向量库通过Embedding模型将文本转化为高维向量，再利用近似最近邻算法（ANN，如HNSW、IVF-Flat）检索语义相近内容67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/899224API
curl https://kongchang.com/api/v1/knowledge/claims/899224MCP
get_claim(id=899224)