Unverified50% confidenceOpinionExact time
Data quality is not equal to cleanliness, but to containing enough effective signal; a dataset that contains noise but has broad coverage is often superior to a dataset that is clean but narrow.
1
Sources
50%
Confidence
Long-term
Relevance
10/15/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe key metric for measuring retrieval quality should be the proportion of effective context, not expanding recall volume, as the latter may introduce noise that dilutes the context budget.72% similarUnverified在 SFT 中少而精的高质量数据往往优于海量的低质量数据69% similarUnverifiedPaired comparisons should be used and confidence intervals reported; small gains that fall within run-to-run noise do not constitute improvement.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1038958API
curl https://kongchang.com/api/v1/knowledge/claims/1038958MCP
get_claim(id=1038958)