Aggregate LLM Evaluation on (MATH, GSM8K, H_Eval, MMLU, CEVAL, CMMLU, BoolQ, CSQA)
81.41Aggregate AccuracyQwen3-30B-A3B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Qwen3-30B-A3BSparsity Level=Base2026.05 | 81.41 | 8 | |
| BEAMSparsity Level=Mid Sparsity, beta=0.012026.05 | 79.99 | 4.23 | |
| Top-K ReducedSparsity Level=Mid Sparsity, K=42026.05 | 78.9 | 4 | |
| MoE-DynamicSparsity Level=Mid Sparsity, phi=0.32026.05 | 77.49 | 5.04 | |
| BEAMSparsity Level=High Sparsity, beta=0.12026.05 | 77.14 | 1.23 | |
| Top-K ReducedSparsity Level=High Sparsity, K=22026.05 | 72.18 | 2 | |
| BEAMSparsity Level=Extreme Sparsity, beta=1.02026.05 | 71.91 | 0.56 | |
| Top-K PruningSparsity Level=Mid Sparsity, K=42026.05 | 69.6 | 4 | |
| MoE-DynamicSparsity Level=High Sparsity, phi=0.12026.05 | 67.93 | 1.74 | |
| AdaMoESparsity Level=Mid Sparsity, Null=1282026.05 | 66.81 | 4.02 | |
| DeepSeekV2-LiteK=6, Avg. K=6.002026.05 | 55.15 | — | |
| BEAMSparsity Level=Mid Sparsity, beta=0.01, Avg. K=2.612026.05 | 55.06 | — | |
| Top-K ReducedSparsity Level=Mid Sparsity, K=4, Avg. K=4.002026.05 | 54.32 | — | |
| Top-K ReducedSparsity Level=Extreme Sparsity, K=12026.05 | 53.95 | 1 | |
| BEAMSparsity Level=High Sparsity, beta=0.1, Avg. K=1.082026.05 | 53.32 | — | |
| AdaMoESparsity Level=High Sparsity, Null=2562026.05 | 51.01 | 2.64 | |
| Top-K PruningSparsity Level=Mid Sparsity, K=4, Avg. K=4.002026.05 | 50.88 | — | |
| Top-K ReducedSparsity Level=High Sparsity, K=2, Avg. K=2.002026.05 | 50.27 | — | |
| MoE-DynamicSparsity Level=Mid Sparsity, phi=0.3, Avg. K=4.312026.05 | 47.7 | — | |
| BEAMSparsity Level=Extreme Sparsity, beta=1.0, Avg. K=0.482026.05 | 47.39 | — | |
| AdaMoESparsity Level=Mid Sparsity, Null=64, Avg. K=3.252026.05 | 47.09 | — | |
| MoE-DynamicSparsity Level=High Sparsity, phi=0.1, Avg. K=3.902026.05 | 44.9 | — | |
| Top-K PruningSparsity Level=High Sparsity, K=2, Avg. K=2.002026.05 | 42.1 | — | |
| AdaMoESparsity Level=High Sparsity, Null=128, Avg. K=2.112026.05 | 41.31 | — | |
| Top-K ReducedSparsity Level=Extreme Sparsity, K=1, Avg. K=1.002026.05 | 35.77 | — | |
| Top-K PruningSparsity Level=High Sparsity, K=22026.05 | 11.92 | 2 |