List your company

Indirect Object Identification on IOI (CPR, CMD, Avg Faithfulness)

3.491CPREAP-IG
0.125560.999281.8732.74672Jun 5, 2026

Evaluation results

MethodLinks
EAP-IGModel=Gemma-2-2B, Path=inputs2026.063.4912.4962.592
EAP-IGModel=Llama-3.1-8B, Path=inputs2026.062.3911.3922.404
EAP-IGModel=GPT-2, Path=inputs2026.062.1551.1591.796
HVPModel=Gemma-2-2B2026.061.910.9271.125
EAP-IGModel=Qwen2.5-0.5B, Path=inputs2026.061.680.6821.537
MS-HVPModel=Llama-3.1-8B, K=52026.061.3760.3791.129
EAPModel=Gemma-2-2B2026.061.3640.380.9
MS-HVPModel=Gemma-2-2B, K=52026.061.3060.3270.814
HVPModel=GPT-22026.061.2810.2920.888
EAPModel=GPT-22026.061.2670.2780.885
MS-HVPModel=GPT-2, K=52026.061.2530.2650.871
EAPModel=Llama-3.1-8B2026.060.9690.0310.767
HVPModel=Llama-3.1-8B2026.060.9390.060.75
HVPModel=Qwen2.5-0.5B2026.060.2640.7350.107
EAPModel=Qwen2.5-0.5B2026.060.2630.7360.107
MS-HVPModel=Qwen2.5-0.5B, K=52026.060.2550.7440.102