Expert Knowledge Q&A on GPQA
42.86AccuracyDense
Evaluation results
| Method | Links | |
|---|---|---|
| DenseModel=Qwen-3 8B, Sparsity=02025.10 | 42.86 | |
| DenseModel=Qwen-3 14B, Sparsity=02025.10 | 39.29 | |
| ARMORModel=Qwen-3 8B, Sparsity=2:4+5.03%2025.10 | 33.93 | |
| SparseGPTModel=Qwen-3 8B, Sparsity=2:42025.10 | 32.14 | |
| ARMORModel=Qwen-3 14B, Sparsity=2:4+3.89%2025.10 | 29.91 | |
| WandaModel=Qwen-3 8B, Sparsity=2:42025.10 | 29.46 | |
| WandaModel=Qwen-3 14B, Sparsity=2:42025.10 | 29.02 | |
| NoWag-PModel=Qwen-3 8B, Sparsity=2:42025.10 | 28.57 | |
| SparseGPTModel=Qwen-3 14B, Sparsity=2:42025.10 | 27.9 | |
| NoWag-PModel=Qwen-3 14B, Sparsity=2:42025.10 | 27.46 |