Reasoning on OCNLI
72.75ScoreLLaDA2.1-flash
Evaluation results
| Method | Links | ||
|---|---|---|---|
| LLaDA2.1-flashInference Mode=Q Mode2026.02 | 72.75 | 1.32 | |
| LLaDA2.1-flashInference Mode=S Mode2026.02 | 72.17 | 1.83 | |
| LLaDA2.0-flash2026.02 | 71.63 | 1.09 | |
| Qwen3-30B-A3B-Inst-25072026.02 | 71.59 | 1 | |
| Ling-flash-2.02026.02 | 65.36 | 1 | |
| LLaDA2.0-mini2026.02 | 64.51 | 4.06 | |
| LLaDA2.1-minimode=Q Mode2026.02 | 61.59 | 1.23 | |
| Qwen3-8Bno think=true2026.02 | 61.49 | — | |
| LLaDA2.1-minimode=S Mode2026.02 | 61.02 | 1.78 | |
| Ling-mini-2.02026.02 | 60.17 | — |