List your company

Commonsense Question Answering on PIQA (Accuracy)

79.49PIQA AccuracyLlama-3.1-8B-Instruct
48.872456.821264.7772.7188Apr 21, 2026Apr 27, 2026May 4, 2026May 11, 2026May 18, 2026May 25, 2026Jun 1, 2026

Evaluation results

MethodLinks
Llama-3.1-8B-InstructModel=Llama-3.1-8B-Instruct, Sparsity=Dense2026.0679.49
DeepSeek-7B-chatModel=DeepSeek-7B-chat, Sparsity=Dense2026.0677.64
FP16Bits=-, Backbone=LLAMA-2-7B, Zero-shot=true2026.0476.93
Llama-3.2-3B-InstructModel=Llama-3.2-3B-Instruct, Sparsity=Dense2026.0675.63
Qwen2.5-7B-InstructModel=Qwen2.5-7B-Instruct, Sparsity=Dense2026.0673.5
LBLLMBits=W(1+1)A4, Backbone=LLAMA-2-7B, Zero-shot=true2026.0470.97
Qwen3-4B-InstructModel=Qwen3-4B-Instruct, Sparsity=Dense2026.0669.7
SUBFITModel=Llama-3.1-8B-Instruct, Sparsity=25%2026.0669.48
SUBFITModel=Qwen2.5-7B-Instruct, Sparsity=25%2026.0669.26
BWABits=W(1+1)A4, Backbone=LLAMA-2-7B, Zero-shot=true2026.0468.72
CBQBits=W4A4, Backbone=LLAMA-2-7B, Zero-shot=true2026.0468.25
SUBFITModel=Llama-3.2-3B-Instruct, Sparsity=25%2026.0665.51
SUBFITModel=DeepSeek-7B-chat, Sparsity=25%2026.0665.02
ARB-LLMBits=W(1+1)A4, Backbone=LLAMA-2-7B, Zero-shot=true2026.0463.82
SUBFITModel=Qwen3-4B-Instruct, Sparsity=25%2026.0655.01
QuaRotBits=W2A4, Backbone=LLAMA-2-7B, Zero-shot=true2026.0454.41
BiLLMBits=W(1+1)A4, Backbone=LLAMA-2-7B, Zero-shot=true2026.0450.05