Using LLM-as-a-Judge grading instead of heuristic answer matching dramatically reduces measured prompt sensitivity and stabilizes model rankings across prompt templates, with human annotations confirming the judge's view.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Flaw or Artifact? Rethinking Prompt Sensitivity in Evaluating LLMs
Using LLM-as-a-Judge grading instead of heuristic answer matching dramatically reduces measured prompt sensitivity and stabilizes model rankings across prompt templates, with human annotations confirming the judge's view.