引导式测验 · 硬 等级
Calibrating LLM Judges Against Humans Quiz
Calibrate LLM judges with shared rubrics, independent human labels, confusion matrices, suitable agreement statistics, and repeated held-out checks.
问题 1 的 8
引导式测验 · 硬 等级
Calibrate LLM judges with shared rubrics, independent human labels, confusion matrices, suitable agreement statistics, and repeated held-out checks.