Verified80% confidenceFactTime unknown
Embedding models typically map text into high-dimensional vector spaces of 768 or 1536 dimensions.
20
Sources
80%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedEmbedding models typically compress text into dense vectors of 768 or 1536 dimensions.77% similarUnverified现代Embedding模型通常基于BERT或其变体架构，输出768维至4096维不等的稠密向量74% similarUnverifiedEmbedding vector dimensions typically range from 768 to 3072; higher dimensions provide stronger expressive capability but at greater computational cost.72% similarUnverifiedtext-embedding-3-small的向量维度为1536维，BGE-small为384维71% similarUnverified嵌入向量通常为768维或1536维70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/26767API
curl https://kongchang.com/api/v1/knowledge/claims/26767MCP
get_claim(id=26767)