List your company

Commonsense Reasoning on OpenBookQA (Inter/Intra-Layer Filtering)

75.6Accuracy (Inter-layer)CRFT
49.656.3563.169.85Jul 14, 2025

Evaluation results

MethodLinks
CRFTBackbone=LLaMA-3-8B, Identify=SSF2025.0775.677
CRFTBackbone=LLaMA-3-8B, Identify=MAF2025.0773.872.6
CRFTBackbone=Mistral-7B, Identify=MSF2025.0772.868.6
CRFTBackbone=Mistral-7B, Identify=SAF2025.0772.672.6
CRFTBackbone=LLaMA-3-8B, Identify=SAF2025.077173.6
CRFTBackbone=LLaMA-3-8B, Identify=MSF2025.077068.6
ReFTBackbone=LLaMA-3-8B, Identify=p7+s72025.076666
CRFTBackbone=Mistral-7B, Identify=MAF2025.0764.269.6
ReFTBackbone=Mistral-7B, Identify=p7+s72025.0763.863.8
CRFTBackbone=Mistral-7B, Identify=SSF2025.0761.666.4
CRFTBackbone=LLaMA-2-7B, Identify=MSF2025.075956.4
ReFTBackbone=LLaMA-2-7B, Identify=p7+s72025.0758.658.6
CRFTBackbone=LLaMA-2-7B, Identify=SSF2025.0758.458.6
CRFTBackbone=LLaMA-2-7B, Identify=SAF2025.075762.2
CRFTBackbone=LLaMA-2-7B, Identify=MAF2025.0750.666.4