Unverified50% confidenceFactExact time
In 13% to 38% of correct answers, the model's reasons for eliminating distractors were clinically untenable, and confidence did not reflect this at all.
1
Sources
50%
Confidence
Long-term
Relevance
10/23/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedWhen a predictive model performs excellently but clinicians cannot take practical action based on it, its impact on outcomes may be minimal.69% similarUnverifiedIn the experiment, the proportion of unhealthy samples was 41.7%, that is, 5 out of 12 samples were unhealthy.69% similarUnverifiedEven the best-performing models can only recover 30% of the evaluated evidence and 25% of the limitations descriptions.69% similarUnverifiedThe model has an 89% probability of finding the correct evidence, but only a 20% probability of turning it into a verifiable product assertion; 71% of features ultimately had no verified capability description, and the workflow recovery rate was only 18%.68% similarVerifiedIn scenarios with extreme class imbalance, even an accuracy as high as 95% may indicate that the model is completely unable to identify the minority class.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1088177API
curl https://kongchang.com/api/v1/knowledge/claims/1088177MCP
get_claim(id=1088177)