List your company

Logical Reasoning on DetectiveQA

88.31Accuracy (DetectiveQA)CDS
70.078874.811979.54584.2781May 13, 2026

Evaluation results

MethodLinks
CDSModel=gpt-5.2, Number of demonstrations=1242026.0588.31
CDSbgeModel=gpt-5.2, Number of demonstrations=1242026.0588.31
CDSbgeModel=gpt-5.2, Number of demonstrations=642026.0586.36
originModel=gpt-5.2, Number of demonstrations=1242026.0585.71
CDSbgeModel=gpt-5.2, Number of demonstrations=322026.0585.71
CDSModel=gpt-5.2, Number of demonstrations=642026.0585.06
originModel=gpt-5.2, Number of demonstrations=642026.0583.77
CDSModel=gpt-5.2, Number of demonstrations=322026.0583.12
originModel=gpt-5.2, Number of demonstrations=322026.0582.47
CDSbgeModel=gpt-5.2, Number of demonstrations=162026.0581.17
originModel=gpt-5.2, Number of demonstrations=162026.0580.52
CDSModel=gpt-5.2, Number of demonstrations=162026.0580.52
CDSbgeModel=Qwen3, Number of demonstrations=322026.0577.29
CDSbgeModel=Qwen3, Number of demonstrations=1242026.0576.63
CDSModel=Qwen3, Number of demonstrations=162026.0576.62
originModel=Qwen3, Number of demonstrations=162026.0575.97
CDSModel=Qwen3, Number of demonstrations=322026.0575.32
CDSModel=Qwen3, Number of demonstrations=1242026.0575.32
CDSbgeModel=Qwen3, Number of demonstrations=162026.0575.32
CDSbgeModel=Qwen3, Number of demonstrations=642026.0575.32
originModel=Qwen3, Number of demonstrations=322026.0574.03
CDSModel=Qwen3, Number of demonstrations=642026.0573.38
originModel=Qwen3, Number of demonstrations=1242026.0572.73
originModel=Qwen3, Number of demonstrations=642026.0570.78