Diagnosing the Reliability of LLM-as-a-Judge via Item Response Theory 文章

ArXiv CS.AI2026-06-01NEWSen作者: Junhyuk Choi, Sohhyung Park, Chanhee Cho, Hyeonchu Park, Bugeun Kim
查看原文 →
关系图谱

Diagnosing the Reliability of LLM-as-a-Judge via Item Response Theory · 相关技术

相关技术

IRT-GRMLLM judgesGraded Response ModelItem Response TheoryLLM-as-a-Judge