Unverified50% confidenceFactExact time
经过 Attention 处理后得到的向量称为 Contextualized Embedding，能结合上下文让相同 Token 根据语境得到不同表示
1
Sources
50%
Confidence
Long-term
Relevance
7/20/2026
First Seen
Sources
Related Claims
UnverifiedThe larger the context, the more input Tokens consumed70% similarVerifiedEmbedding是将文本转换为高维数值向量的过程，语义相近的文本在向量空间中距离更近68% similarUnverifiedIn semantic space, semantically similar inputs are mapped to positions close to each other, while semantically dissimilar inputs are pushed apart.68% similarUnverifiedVector embedding is the process of transforming discrete text, code, or experiences into numerical vectors in a continuous high-dimensional space, so that semantically similar content is closer in vector space.64% similarUnverified语义缓存通过将用户请求转化为向量嵌入，在向量空间中计算语义相似度，将语义相近的问题映射到同一缓存结果，从而减少模型调用次数和Token消耗64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/570189API
curl https://kongchang.com/api/v1/knowledge/claims/570189MCP
get_claim(id=570189)