How do different LLM judges rank the same papers?

INTERACTIVE RANK AGGREGATION

Select one judge—or combine any two, three, four, or all five.

Each model is normalized to its within-model percentile rank. Combined rankings use the mean or median percentile, so models with different raw score scales receive equal weight.