Unverified50% confidenceCautionExact time
若将验证集或测试集数据纳入 fit，会造成数据泄露，导致模型评估结果过于乐观
1
Sources
50%
Confidence
Long-term
Relevance
7/17/2026
First Seen
Sources
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84% similarUnverifiedSome models carry the risk of over-optimization toward test sets, causing leaderboard scores to deviate from real-world performance (benchmark saturation).79% similarUnverifiedData leakage falls into two categories—target leakage and train-test contamination—and leads to systematic overestimation of model performance.75% similarUnverified当标注不一致存在时，增加更多数据只会成倍放大混乱，向模型提供自相矛盾的监督信号73% similarVerifiedBenchmark data contamination refers to the phenomenon where, after a test set is made public, its questions may appear in the pre-training corpora of subsequent models, causing scores to reflect memorization rather than genuine reasoning ability.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/542338API
curl https://kongchang.com/api/v1/knowledge/claims/542338MCP
get_claim(id=542338)