List your company

Debiasing Effectiveness on In-Distribution (ID)

10.2Mean Effectiveness Score (ID)ORIGINAL
0.50723.02365.548.0564Mar 19, 2026

Evaluation results

MethodLinks
ORIGINALModel Family=QWEN2.5-7B2026.0310.227.43
ORIGINALModel Family=QWEN2.5-14B2026.038.7133.13
ORIGINALModel Family=QWEN2.5-3B2026.037.4120.08
ORIGINALModel Family=GEMMA-2-2B2026.036.8437.45
UGIDModel Family=QWEN2.5-7B2026.031.152.24
CDAModel Family=QWEN2.5-3B2026.031.141.29
CDAModel Family=QWEN2.5-14B2026.031.141.29
CDAModel Family=GEMMA-2-2B2026.031.131.21
CDAModel Family=QWEN2.5-7B2026.031.121.14
KLAADModel Family=QWEN2.5-7B2026.031.081.13
KLAADModel Family=GEMMA-2-2B2026.031.011.06
UGIDModel Family=QWEN2.5-3B2026.030.981
KLAADModel Family=QWEN2.5-14B2026.030.981.13
UGIDModel Family=GEMMA-2-2B2026.030.951
UGIDModel Family=QWEN2.5-14B2026.030.941
KLAADModel Family=QWEN2.5-3B2026.030.880.88