Last released Jul 2, 2026
Evaluate LLM-as-a-Judge systems by measuring alignment between judge outputs and human annotations
Supported by