List your company

Honesty Alignment on HonestyBench OOD

81.19Squad ScoreCal-Only
55.647662.278868.9175.5412Oct 20, 2025

Evaluation results

MethodLinks
Cal-OnlyCategory=Upper Bound, Amount of annotated data=560k, Base Model=Qwen2.5-7B-Instruct2025.1081.1981.380.4579.5892.1185.75
EliCalCategory=Upper Bound, Amount of annotated data=560k, Base Model=Qwen2.5-7B-Instruct2025.1081.0481.181.0280.6892.1185.83
EliCalCategory=Training-based, Amount of annotated data=1k, Base Model=Qwen2.5-7B-Instruct2025.1078.4880.1179.8578.0991.7484.47
Verbal-10Category=Training-free, Base Model=Qwen2.5-7B-Instruct2025.1072.5468.263.2564.4473.471.05
Cal-OnlyCategory=Training-based, Amount of annotated data=1k, Base Model=Qwen2.5-7B-Instruct2025.1071.5971.4869.3366.9686.1377.32
DACACategory=Training-based, Base Model=Qwen2.5-7B-Instruct2025.1066.7965.7162.6273.9876.470.98
Consis-SemCategory=Training-free, Base Model=Qwen2.5-7B-Instruct2025.1066.1676.2677.570.7670.4470.2
Verbal-0Category=Training-free, Base Model=Qwen2.5-7B-Instruct2025.1065.7670.4159.5660.5470.6467.12
Consis-LexCategory=Training-free, Base Model=Qwen2.5-7B-Instruct2025.1062.1265.4372.5961.0777.0769.87
N-ProbCategory=Training-free, Base Model=Qwen2.5-7B-Instruct2025.1060.7266.0670.5165.9374.7368.58
Eli-OnlyCategory=Training-based, Base Model=Qwen2.5-7B-Instruct2025.1060.6676.6174.7766.5674.669.66
ThermometerCategory=Training-based, Base Model=Qwen2.5-7B-Instruct2025.1060.2362.956.9861.968.7663.88
ProbCategory=Training-free, Base Model=Qwen2.5-7B-Instruct2025.1056.6361.368.3461.8571.364.94