mean	llm_as_judge
0.488	0.488