Unverified50% confidenceSolutionExact time
After user confirmation, the confirmed "photo-identity" pairs can be written back as positive samples and the rejected candidates as negative samples for subsequent evaluation or threshold fine-tuning; the written-back data needs to be de-identified.
1
Sources
50%
Confidence
Long-term
Relevance
11/7/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIf the XAI evaluation metrics themselves are untrustworthy, method rankings based on them may be distorted, so metrics that have passed meta-evaluation and show high agreement with other metrics should be preferred.65% similarUnverifiedThe positive control is borrowed from experimental science and means testing a known necessarily positive control sample before testing the sample to prove the detection process works; the negative control confirms that records that should not be returned do not appear; neither carries information about the time dimension.65% similarUnverifiedContext-blind evaluation treats each question-answer segment as an independent sample, checking only whether the reply itself conforms to medical common sense and safety norms; user context-aware evaluation takes the user's previously accumulated historical information such as age, gender, past medical history, and medication status as preconditions, and then judges whether this reply is safe for this specific person.65% similarUnverified低成本标注一致性检查方法:从数据集挑选20-30张困难图像,由两名标注员独立标注64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1146030API
curl https://kongchang.com/api/v1/knowledge/claims/1146030MCP
get_claim(id=1146030)