Unverified50% confidenceFactExact time
Word2Vec训练后向量算术运算能捕捉语义关系，例如 vec(king) - vec(man) + vec(woman) ≈ vec(queen)
1
Sources
50%
Confidence
Long-term
Relevance
8/18/2026
First Seen
Sources
Related Claims
Unverified原始word2vec需要在整个词汇表上计算softmax归一化，计算复杂度为O(V)，负采样将问题转化为二分类以降低计算量64% similarUnverifiedWord2Vec、SVD、GloVe是三种不同的用来学习embedding矩阵的方法63% similarUnverifiedWord2Vec通过CBOW或Skip-gram两种训练方式，将词语压缩为100～300维的稠密向量61% similarUnverifiedsentence-transformers库在BERT基础上引入了孪生网络训练范式，使输出向量在余弦相似度空间中具备良好的语义对齐性60% similarUnverifiedEarly embedding models like Word2Vec and GloVe could only handle word-level representations, while modern models can encode entire sentences or paragraphs60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/771180API
curl https://kongchang.com/api/v1/knowledge/claims/771180MCP
get_claim(id=771180)