Generative Language Tasks on GEN benchmark
93.4IFEvalOriginal
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| OriginalModel=Qwen3-80B-A3B-Instruct, Experts (N)=5122026.04 | 93.4 | 80 | 78.6 | 47 | 95.1 | 43.2 | 72.9 | |
| REAMModel=Qwen3-80B-A3B-Instruct, Experts (N)=3842026.04 | 93.4 | 73.3 | 78.1 | 46.5 | 93.9 | 43.7 | 71.5 | |
| REAPModel=Qwen3-80B-A3B-Instruct, Experts (N)=3842026.04 | 92.8 | 66.7 | 77.7 | 42.4 | 94.5 | 43.6 | 69.6 | |
| OriginalModel=GLM-4.5-Air, Experts (N)=1282026.04 | 90.4 | 83.3 | 94.8 | 42.9 | 93.9 | 57.4 | 77.1 | |
| OriginalModel=Qwen3-Coder-Next, Experts (N)=5122026.04 | 89.6 | 80 | 85.4 | 42.4 | 92.7 | 47.5 | 72.9 | |
| REAMModel=Qwen3-Coder-Next, Experts (N)=3842026.04 | 89.3 | 80 | 85.3 | 40.4 | 94.5 | 48 | 72.9 | |
| REAPModel=Qwen3-Coder-Next, Experts (N)=3842026.04 | 87.5 | 70 | 86.4 | 37.9 | 94.5 | 47.7 | 70.7 | |
| REAMModel=GLM-4.5-Air, Experts (N)=962026.04 | 83.6 | 83.3 | 94.9 | 37.9 | 90.2 | 53.7 | 73.9 | |
| REAPModel=GLM-4.5-Air, Experts (N)=962026.04 | 80.6 | 76.7 | 93.9 | 38.4 | 90.2 | 51.7 | 71.9 |