Logical and Commonsense Reasoning on Reasoning Suite
89.36BIG-Bench HardLing-flash-2.0
Evaluation results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Ling-flash-2.02025.12 | 89.36 | 23.24 | 75.09 | 82.72 | 87.6 | 97.88 | 91.95 | 65.36 | 81.59 | 68.96 | 88.32 | 81.32 | |
| LLaDA2.0-flash2025.12 | 86.75 | 27.86 | 87.52 | 80.48 | 82.3 | 96.5 | 92.76 | 71.63 | 84.97 | 64.24 | 87.9 | 90 | |
| Qwen3-30B-A3B-Instruct-25072025.12 | 85.54 | 37.8 | 86.18 | 79.15 | 90.97 | 97.12 | 91.57 | 71.59 | 86.31 | 68 | 87.57 | 89.51 | |
| LLaDA2.0-flash-preview2025.12 | 82.85 | 16.7 | 83.38 | 78.75 | 39.9 | 93.5 | 91.84 | 69.39 | 86 | 53.28 | 88.17 | 90.61 |