Unverified50% confidenceBenchmarkExact time
Before adding the GIN index and the generated tsvector, searching across 10,000 translations took about 800ms; after adding the index, even when the data volume grew to 100,000 translations, the same query dropped to about 20-40ms.
1
Sources
50%
Confidence
Long-term
Relevance
10/14/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedDividing 200 tools into 10 namespaces with 20 tools per group can reduce the search scope from 200 to 20.60% similarUnverifiedBERT fine-tuning typically requires only hundreds to thousands of samples to achieve good results on downstream classification tasks.54% similarUnverifiedIn the Indian number naming system, 1 lakh equals 100,000 and 1 crore equals 10,000,000, written using a two-digit grouping convention.52% similarUnverifiedEnglish Wikipedia has more than 6.5 million articles, while some language editions have only hundreds of thousands.51% similarUnverifiedThe number of SQLite deployments is estimated to exceed 1 trillion instances.51% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1023226API
curl https://kongchang.com/api/v1/knowledge/claims/1023226MCP
get_claim(id=1023226)