Unverified60% confidenceFactExact time
One-hot encoding represents each word as a sparse vector of vocabulary-size dimensions, and the cosine similarity between any two one-hot vectors is always 0.
2
Sources
60%
Confidence
Long-term
Relevance
7/23/2026
First Seen
Sources
词向量到大语言模型：AI 2.0技术演进全解析
Related Claims
UnverifiedChen Jingrun proved '1+2', namely that any sufficiently large even number N can either be written as the sum of two primes, or as the sum of a prime and the product of two primes.60% similarUnverified1M 上下文意味着模型可以一次性处理约 75 万个英文单词或数十万行代码59% similarUnverifiedSynonym normalization can be achieved by using sentence embedding models such as Sentence-BERT to map words into dense vectors, and then computing semantic proximity via cosine similarity.59% similarVerified余弦相似度是衡量两个向量方向接近程度的指标，值域在-1到1之间，越接近1表示语义越相似59% similarVerifiedSemantic retrieval is based on mathematical metrics such as cosine similarity rather than simple keyword matching.58% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/600454API
curl https://kongchang.com/api/v1/knowledge/claims/600454MCP
get_claim(id=600454)