Unverified50% confidenceFactExact time
Saturation in measurement theory refers to test scores losing the ability to discriminate between the subjects being measured; when all models' scores approach the ceiling or floor, the variance is extremely small.
1
Sources
50%
Confidence
Long-term
Relevance
10/13/2026
First Seen
Sources
Related Entities
Related Claims
Verified若用含测试集的全量数据计算均值和标准差，会导致预处理泄漏，使评估指标虚高，无法真实反映模型泛化能力78% similarUnverifiedBenchmark saturation refers to the phenomenon where a benchmark loses its discriminative power when model performance approaches human-level performance or the upper limit of the test set.76% similarUnverified天花板效应指当测试工具的难度上限低于被测对象的实际能力时，分数分布呈现严重负偏态，导致测试的信度和效度显著下降74% similarUnverifiedThe level of probe accuracy reflects the degree of linear separability of the information, not whether the model intentionally represents the concept.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1013424API
curl https://kongchang.com/api/v1/knowledge/claims/1013424MCP
get_claim(id=1013424)