Unverified50% confidenceFactExact time
Cosine similarity measures the angle between two vectors, where a result of 1 means the same direction (the same meaning) and 0 means irrelevant; it ignores vector length, so a short question can also match a long document.
1
Sources
50%
Confidence
Long-term
Relevance
10/13/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedCosine similarity and Euclidean distance may give diametrically opposite judgments about which two vectors are more similar.75% similarUnverified余弦相似度是语义缓存最常用的距离度量，通常将0.92以上视为缓存命中的保守阈值72% similarUnverifiedWilson置信区间相比简单正态近似在小样本场景下更为准确，能有效避免通过率接近0或1时出现区间越界的问题64% similarUnverifiedCommonly used distance metrics for vector retrieval include cosine similarity and dot product; practical systems typically employ approximate nearest neighbor (ANN) index structures such as HNSW or IVF to trade a small amount of precision for speed.63% similarUnverifiedROUGE metrics exhibit length bias: models that generate shorter summaries have an advantage in precision, models that generate longer summaries have an advantage in recall, and ROUGE cannot measure factual consistency.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1014326API
curl https://kongchang.com/api/v1/knowledge/claims/1014326MCP
get_claim(id=1014326)