| Benchmark | Model | Rank | Results |
|---|---|---|---|
| mathematical-reasoning-on-olympiadbench | Qwen3-30B-A3B-Thinking with PUMA | #5 | Accuracy: 72.3 |
| mathematical-reasoning-on-aime24 | Qwen3-30B-A3B-Thinking with PUMA | #12 | Acc: 90.0 |
| mathematical-reasoning-on-math-500 | Qwen3-30B-A3B-Thinking with PUMA | #36 | Accuracy: 94.2 |
| mathematical-reasoning-on-aime-2025 | Qwen3-30B-A3B-Thinking with PUMA | #41 | Accuracy: 80.0 |
| multiple-choice-question-answering-mcqa-on-gpqa-diamond | Qwen3-30B-A3B-Thinking with PUMA | #67 | Accuracy: 75.8 |