Unverified50% confidenceFactExact time
Entry-point contamination is a classic problem in the field of data annotation, referring to bias introduced at the starting point of data generation or collection that is difficult to eliminate later.
1
Sources
50%
Confidence
Long-term
Relevance
10/16/2026
First Seen
Sources
Related Entities
Related Claims
Unverified污点分析通过标记污点源(Source)、汇聚点(Sink)和净化函数(Sanitizer)追踪不可信数据流向以定位注入风险67% similarVerifiedData contamination refers to a situation where a model has already seen data from the test set during training, causing benchmark scores to be inflated.66% similarUnverifiedThe study redefines the benchmark contamination problem as a violation of anchor-item invariance.65% similarUnverifiedNLP基准数据集被广泛使用后会出现数据污染（Data Contamination）现象，导致基准分数虚高63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1041432API
curl https://kongchang.com/api/v1/knowledge/claims/1041432MCP
get_claim(id=1041432)