Unverified50% confidenceBenchmarkExact time
Under the leave-one-author-out cross-validation setting, the strongest model achieved an accuracy of 81.42%.
1
Sources
50%
Confidence
Long-term
Relevance
9/11/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedOn reasoning-based question-answering tasks, the strongest closed-source models achieved accuracy exceeding 90%.71% similarUnverified作者验证有93%的概率质量正确对齐，信息损失并非映射逻辑的bug71% similarUnverifiedThe per-step accuracy of current top-tier models is approximately between 90% and 95%.68% similarUnverifiedIn mainstream Text-to-SQL benchmarks such as Spider and BIRD, top models have already exceeded 85% accuracy.68% similarUnverified研究表明经过良好校准的LLM裁判与人工标注的一致性可达80%以上67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/902500API
curl https://kongchang.com/api/v1/knowledge/claims/902500MCP
get_claim(id=902500)