Unverified50% confidenceFactExact time
Traditional pseudo-labeling suffers from confirmation bias: repeatedly using pseudo-labels from the same model causes biases to accumulate and amplify with each iteration.
1
Sources
50%
Confidence
Long-term
Relevance
9/28/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe widespread mislabeling will dilute the credibility of AI labeling systems, causing users to ignore such labels.70% similarUnverifiedRelying solely on feature matching may be insufficient to identify sophisticated, carefully crafted fake accounts.69% similarUnverified在关系抽取中，标签出现的频率几乎不能预测模型在该标签上的表现，真正决定泛化能力的是模型是否见过相似的证据表达方式69% similarUnverifiedEarly research found that even when using examples with randomly incorrect labels, few-shot performance was sometimes not much worse than with correct labels.68% similarUnverified评估分数依赖测试数据集，数据集随业务变化会导致评分漂移，使历史横向比较产生误导68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/955558API
curl https://kongchang.com/api/v1/knowledge/claims/955558MCP
get_claim(id=955558)