Composite Reasoning Evaluation (BBH, BBH-XH, SQuAD 2.0, DROP, PIQA, etc.)
83.7BBHLing-mini-2.0
Evaluation Results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Ling-mini-2.02025.12 | 83.7 | 14.81 | 66.11 | 71.36 | 79.85 | 96.06 | 87.54 | 60.17 | 69.02 | 62.72 | 78.8 | 75.56 | |
| Qwen3-8Bthinking=no think2025.12 | 79.48 | 18.27 | 80.09 | 70.02 | 37.48 | 93.12 | 88.3 | 61.49 | 79.56 | 54.48 | 84.56 | 85.21 | |
| LLaDA2.0-mini2025.12 | 78.21 | 16.47 | 75.75 | 71.48 | 64.2 | 86 | 86.51 | 64.51 | 79.01 | 50.4 | 81.91 | 86.5 | |
| LLaDA2.0-miniversion=preview2025.12 | 70.64 | 12.36 | 66.62 | 56.77 | 14.8 | 70 | 84.33 | 58.68 | 74.01 | 37.26 | 79.49 | 85.61 |