List your company

Zero-shot Evaluation on ARC (c/e), OBQA, WinoGrande, PIQA, HellaSwag, MathQA, RTE, and BoolQ

71.08Mean AccuracyOriginal (Dense)
44.24851.21458.1865.146Mar 9, 2026

Evaluation results

MethodLinks
Original (Dense)Model=Qwen3-30B-A3B, Sparsity Ratio=20%2026.0371.08
Original (Dense)Model=Qwen3-30B-A3B, Sparsity Ratio=40%2026.0371.08
Original (Dense)Model=Qwen3-30B-A3B, Sparsity Ratio=60%2026.0371.08
OursModel=Qwen3-30B-A3B, Sparsity Ratio=20%2026.0370.04
NAEEModel=Qwen3-30B-A3B, Sparsity Ratio=20%2026.0369.94
Camera-PModel=Qwen3-30B-A3B, Sparsity Ratio=20%2026.0369.94
Original (Dense)Model=LLaMA-13B, Sparsity Ratio=20%2026.0368.79
Original (Dense)Model=LLaMA-13B, Sparsity Ratio=50%2026.0368.79
OursModel=Qwen3-30B-A3B, Sparsity Ratio=40%2026.0367.65
Camera-PModel=Qwen3-30B-A3B, Sparsity Ratio=40%2026.0367.46
DDPModel=LLaMA-13B, Sparsity Ratio=20%2026.0366.89
Original (Dense)Model=LLaMA-2-7B, Sparsity Ratio=20%2026.0366.63
Original (Dense)Model=LLaMA-2-7B, Sparsity Ratio=50%2026.0366.63
D2-MoEModel=Qwen3-30B-A3B, Sparsity Ratio=20%2026.0366.35
Original (Dense)Model=LLaMA-7B, Sparsity Ratio=20%2026.0365.96
Original (Dense)Model=LLaMA-7B, Sparsity Ratio=50%2026.0365.96
NAEEModel=Qwen3-30B-A3B, Sparsity Ratio=40%2026.0365.77
DDPModel=LLaMA-2-7B, Sparsity Ratio=20%2026.0364.82
LoRAPModel=LLaMA-13B, Sparsity Ratio=20%2026.0364.71
SlimLLMModel=LLaMA-13B, Sparsity Ratio=20%2026.0364.21
DDPModel=LLaMA-7B, Sparsity Ratio=20%2026.0364.13
OursModel=Qwen3-30B-A3B, Sparsity Ratio=60%2026.0363.35
D2-MoEModel=Qwen3-30B-A3B, Sparsity Ratio=40%2026.0362.77
SlimLLMModel=LLaMA-7B, Sparsity Ratio=20%2026.0362.41
Original (Dense)Model=DeepSeekMoE-16B-Base, Sparsity Ratio=20%2026.0362.28
Original (Dense)Model=DeepSeekMoE-16B-Base, Sparsity Ratio=40%2026.0362.28
Original (Dense)Model=DeepSeekMoE-16B-Base, Sparsity Ratio=60%2026.0362.28
DDPModel=LLaMA-13B, Sparsity Ratio=50%2026.0362.14
OursModel=DeepSeekMoE-16B-Base, Sparsity Ratio=20%2026.0361.84
LoRAPModel=LLaMA-7B, Sparsity Ratio=20%2026.0361.7
SlimLLMModel=LLaMA-2-7B, Sparsity Ratio=20%2026.0361.7
LoRAPModel=LLaMA-2-7B, Sparsity Ratio=20%2026.0361.2
Camera-PModel=DeepSeekMoE-16B-Base, Sparsity Ratio=20%2026.0361.03
HEAPrModel=DeepSeekMoE-16B-Base, Sparsity Ratio=20%2026.0360.62
NAEEModel=DeepSeekMoE-16B-Base, Sparsity Ratio=20%2026.0360.51
OursModel=DeepSeekMoE-16B-Base, Sparsity Ratio=40%2026.0360.38
LoRAPruneModel=LLaMA-7B, Sparsity Ratio=20%2026.0360.05
Camera-PModel=Qwen3-30B-A3B, Sparsity Ratio=60%2026.0359.03
D2-MoEModel=DeepSeekMoE-16B-Base, Sparsity Ratio=20%2026.0358.97
Camera-PModel=DeepSeekMoE-16B-Base, Sparsity Ratio=40%2026.0358.58
LoRAPModel=LLaMA-13B, Sparsity Ratio=50%2026.0358.2
OursModel=DeepSeekMoE-16B-Base, Sparsity Ratio=60%2026.0358.18
HEAPrModel=DeepSeekMoE-16B-Base, Sparsity Ratio=40%2026.0357.7
NAEEModel=Qwen3-30B-A3B, Sparsity Ratio=60%2026.0357.13
DDPModel=LLaMA-2-7B, Sparsity Ratio=50%2026.0356.7
DDPModel=LLaMA-7B, Sparsity Ratio=50%2026.0356.07
NAEEModel=DeepSeekMoE-16B-Base, Sparsity Ratio=40%2026.0354.94
SlimLLMModel=LLaMA-13B, Sparsity Ratio=50%2026.0354.75
D2-MoEModel=DeepSeekMoE-16B-Base, Sparsity Ratio=40%2026.0354.32
SlimLLMModel=LLaMA-7B, Sparsity Ratio=50%2026.0353.16
LoRAPModel=LLaMA-2-7B, Sparsity Ratio=50%2026.0352.31
LoRAPModel=LLaMA-7B, Sparsity Ratio=50%2026.0352.17
D2-MoEModel=Qwen3-30B-A3B, Sparsity Ratio=60%2026.0352.02
SlimLLMModel=LLaMA-2-7B, Sparsity Ratio=50%2026.0352.02
Camera-PModel=DeepSeekMoE-16B-Base, Sparsity Ratio=60%2026.0351.62
HEAPrModel=DeepSeekMoE-16B-Base, Sparsity Ratio=60%2026.0351.38
LoRAPruneModel=LLaMA-7B, Sparsity Ratio=50%2026.0349.71
D2-MoEModel=DeepSeekMoE-16B-Base, Sparsity Ratio=60%2026.0346.72
NAEEModel=DeepSeekMoE-16B-Base, Sparsity Ratio=60%2026.0345.28