Primary sourceJerry Liu
View sourceEvent ·Single source
LlamaIndex 用置信度分数决定文档提取自动化阈值
LlamaIndex 发布博客,讨论如何用置信度分数控制文档提取的自动化程度。文章围绕置信度 cutoff、precision 与 recall、score coverage 和 score granularity 展开,说明阈值调高时自动化比例下降但准确率上升。团队用 ExtractBench 对比不同提取系统在置信度过滤后的表现,在 97% precision 目标下,LlamaParse Agentic Plus 过滤后达到 66.48% recall。
Current read
Jerry Liu 讲了置信度分数怎么用:阈值高少自动但更准,低了反之,还拿 ExtractBench 跑了实测数据。
2 sources73° AI heatLast updated 2026/9/24 23:33:37
Evidence
2 sourcesRelated source 1LlamaIndex
View sourceConflict check
Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.