GSM8K gsm8k Leaderboard
Auto-discovered from papers reporting GSM8K (Acc). Β· Metric: Acc (higher is better) Β· π’ Updated 1h ago
| # | Model | Acc | Paper |
|---|---|---|---|
| 1 | LLaDA2.1-Mini SingleBD (Native) | 93.03 | link |
| 2 | LLaDA2.1-Mini MultiBD (training-free) | 92.27 | link |
| 3 | LLaDA2-Mini-CAP SingleBD (Native) | 91.74 | link |
| 4 | LLaDA2-Mini-CAP MultiBD (training-free) | 91.21 | link |
| 5 | Base β (original) | 45.94 | link |
| 6 | Base OPM + XKD z (ours) | 44.70 | link |
| 7 | Base OPM + XKD h (ours) | 44.12 | link |
| 8 | Chat OPM + XKD z (ours) | 41.85 | link |
| 9 | Chat β (original) | 41.32 | link |
| 10 | Chat OPM + XKD h (ours) | 40.41 | link |
| 11 | Base OPM + UT | 36.58 | link |
| 12 | Chat OPM + UT | 33.59 | link |
| 13 | Chat OPM | 3.26 | link |
| 14 | Base OPM | 1.52 | link |