List your company

Inter-Rater Reliability (Kappa/ICC) on MT-Bench

71.88KappaCalibraEval
12.059227.589643.1258.6504Oct 20, 2024

Evaluation results

MethodLinks
CalibraEvalBase Model=Qwen-72B2024.1071.8895.796.71
Qwen-72B (Default)Debiasing=None2024.1071.3590.3391.16
CalibraEvalBase Model=Llama-3-8B2024.1028.6375.4576.8
PrideBase Model=Llama-3-8B2024.1027.3872.2774.5
DIBase Model=Llama-3-8B2024.1015.9359.1165
Llama-3-8B (Default)Debiasing=None2024.1014.3660.9673.08