Study Finds LLM Judge Verdicts Vary With Rubric Wording in Korean Support Dialogues
A new arXiv paper tests whether LLM judges evaluating Korean support dialogues give different verdicts depending on how evaluation rubrics are written in Korean, finding that rubric language and clause structure affect outcomes.
Read the full story at arXiv cs.CL ↗
Timeline · 1 report
- 2026-11-04 05:00 · arXiv cs.CL
Same Rule, Different Verdicts: Rubric Language and Clause Structure in LLM Judges for Korean Support Dialogues