Unverified50% confidenceDecision RuleExact time
The semantic cache uses a cosine similarity threshold of 0.92 as the criterion for determining a cache hit.
1
Sources
50%
Confidence
Long-term
Relevance
10/11/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedA cosine similarity threshold is reliable only if it has been measured on your own corpus with the exact model in use, and re-measured whenever either changes.69% similarUnverifiedPrompt caching is based on prefix matching, reusing already-computed KV Cache to skip redundant computation, so fixed content at the beginning of prompts should be kept consistent to maximize cache hit rate.68% similarUnverifiedCommon implementation approaches for LLM semantic-level assertions include rule-based checkers, embedding vector cosine similarity thresholds (e.g., 0.85), LLM-as-judge scoring, and NLI model-based factual consistency verification.63% similarUnverifiedThe developer added a similarity threshold check before LLM processing, so that only retrieval results with similarity exceeding the set threshold are passed to the LLM.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/998228API
curl https://kongchang.com/api/v1/knowledge/claims/998228MCP
get_claim(id=998228)