List your company

Confidence Estimation on AlpacaEval

0.4395Rank Correlation (RK)Verbalized Confidence
0.27830.320150.3620.40385May 14, 2026

Evaluation results

MethodLinks
Verbalized ConfidenceModel=Llama3-7B2026.050.43950.5589
Predictive ProbabilityModel=Llama3-7B2026.050.43390.557
Random AnnotatorModel=Llama3-7B2026.050.42690.574
Predictive ProbabilityModel=Qwen2.5-32B2026.050.41590.585
Simulated AnnotatorsModel=Llama3-7B2026.050.41290.5904
Random AnnotatorModel=Qwen2.5-32B2026.050.39580.6022
Simulated AnnotatorsModel=Qwen2.5-32B2026.050.38740.6107
Learning Confidence (Vanilla)Model=Llama3-7B2026.050.38710.6189
Learning Confidence (Vanilla)Model=Qwen2.5-32B2026.050.33450.6679
Learning Confidence (Ours)Model=Llama3-7B2026.050.32970.671
Learning Confidence (Ours)Model=Qwen2.5-32B2026.050.28450.7012