← back
arXivIlija Subasic, Andrew Rabinovich, Zhao ChenFri, Oct 30, 2026, 4:02 PM PDT
score 14.9

Method Aligns AI Judge Scores With Human Ratings

Original: Quantiles, Not Points: Streaming Quantile Matching for Ordinal-Aware LLM-as-a-Judge Calibration

Source: arxiv.org ↗

Writing ELI5 summary…