List your company

Commonsense Reasoning on Commonsense Tasks

81.35BoolQ AccuracyDense
36.05847.816559.57571.3335May 22, 2025

Evaluation results

MethodLinks
DenseBudget=100%, Base Model=Llama3-8B, Zero-shot=true2025.0581.3580.7872.6179.1477.6953.334569.99
DenseBudget=100%, Base Model=Llama2-7B, Zero-shot=true2025.0577.7479.1168.9775.9874.6246.2544.266.7
RAPBudget=80%, Base Model=Llama3-8B, Zero-shot=true2025.0568.8476.5566.1165.5562.1239.5139.659.77
RAPBudget=80%, Base Model=Llama2-7B, Zero-shot=true2025.0562.8173.9963.3865.7760.3536.6936.657.08
ShortGPTBudget=80%, Base Model=Llama2-7B, Zero-shot=true2025.0562.1760.1260.3843.741.2530.123547.53
LLMPrunerBudget=80%, Base Model=Llama3-8B, Zero-shot=true2025.0562.1765.0251.9342.3241.3325.529.445.37
SliceGPTBudget=80%, Base Model=Llama2-7B, Zero-shot=true2025.0561.9965.2958.0843.4352.2732.0828.9948.88
SliceGPTBudget=80%, Base Model=Llama3-8B, Zero-shot=true2025.0561.8765.9454.0542.5154.3731.0627.848.23
SliceGPTBudget=60%, Base Model=Llama2-7B, Zero-shot=true2025.0561.0454.5649.4928.9930.6823.4625.439.09
RAPBudget=60%, Base Model=Llama2-7B, Zero-shot=true2025.0557.1658.2653.7537.8137.7926.0230.243
ShortGPTBudget=80%, Base Model=Llama3-8B, Zero-shot=true2025.0556.8258.5954.8537.7136.9930.22843.3
LLMPrunerBudget=60%, Base Model=Llama2-7B, Zero-shot=true2025.0555.9661.1550.8338.0935.0627.1328.442.38
FFN-SkipBudget=80%, Base Model=Llama3-8B, Zero-shot=true2025.0555.7751.4750.7425.925.7525.252837.56
ShortGPTBudget=60%, Base Model=Llama2-7B, Zero-shot=true2025.0555.5750.9850.5127.8326.3927.4726.637.91
RAPBudget=60%, Base Model=Llama3-8B, Zero-shot=true2025.0552.256.6950.3631.9133.5424.2327.439.48
FFN-SkipBudget=60%, Base Model=Llama3-8B, Zero-shot=true2025.0551.9351.9550.0326.0224.0326.9628.5937.08
LLMPrunerBudget=80%, Base Model=Llama2-7B, Zero-shot=true2025.0550.6368.8254.1452.6649.3330.234.5948.63
MHA-DropBudget=80%, Base Model=Llama2-7B, Zero-shot=true2025.0546.0854.5650.8329.1228.6227.4725.237.41
FFN-SkipBudget=60%, Base Model=Llama2-7B, Zero-shot=true2025.0544.8350.6448.625.8625.7228.9228.1936.13
FFN-SkipBudget=80%, Base Model=Llama2-7B, Zero-shot=true2025.0542.849.3551.2226.6927.1828.4926.636.05
ShortGPTBudget=60%, Base Model=Llama3-8B, Zero-shot=true2025.0541.4450.9850.0326.7425.4625.3426.835.26
SliceGPTBudget=60%, Base Model=Llama3-8B, Zero-shot=true2025.0540.8653.8648.9327.9732.0723.0425.836.08
LLMPrunerBudget=60%, Base Model=Llama3-8B, Zero-shot=true2025.0540.4850.0552.3326.2626.7725.942735.55
MHA-DropBudget=60%, Base Model=Llama2-7B, Zero-shot=true2025.0537.8349.8949.8825.7226.0525.4326.834.51
MHA-DropBudget=60%, Base Model=Llama3-8B, Zero-shot=true2025.0537.8351.950.1226.0427.8622.9525.834.64
MHA-DropBudget=80%, Base Model=Llama3-8B, Zero-shot=true2025.0537.853.9250.226.8730.4723.8127.235.75