LLM Judge Reliability

Papers examining biases, artifacts, and consistency issues when using large language models as evaluators for text quality, ranking decisions, and comparative judgments.