Unverified50% confidenceBenchmarkExact time
When the incorrect answer key appeared in the prompt, the model's responses matched that incorrect key in 63% (47/75) of cases.
1
Sources
50%
Confidence
Long-term
Relevance
10/2/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn the control group where the answer key was removed from the prompt, the match rate plummeted from 63% to 1% (1/75).70% similarUnverifiedA single "trusted source" prompt caused 7 out of 8 tested models to reverse between 45% and 88% of their correct answers.64% similarUnverifiedThe traditional expert codebook revision approach achieved an LLM annotation accuracy of only 57.8%.63% similarUnverified当综合评估指数中顶尖模型得分接近天花板时会出现基准饱和问题，60分对59分的差距在统计学上通常处于误差范围之内63% similarUnverifiedOn Track B, the flat RAG baseline can detect 76%–97% of true contradictions, but produces false positives on 16%–43% of surface-similar safe drafts (the specific ratio depends on the backend).62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/968641API
curl https://kongchang.com/api/v1/knowledge/claims/968641MCP
get_claim(id=968641)