| Benchmark | Model | Rank | Results |
|---|---|---|---|
| multi-task-language-understanding-on-mmlu-pro | Nemotron 3 Super (post-trained) | #28 | Accuracy: 83.73 |
| code-generation-on-humaneval | Nemotron 3 Super (120B-A12B Base) | #61 | Pass@1: 79.4 |
| arithmetic-reasoning-on-gsm8k | Nemotron 3 Super (120B-A12B Base) | #67 | Accuracy: 90.67 |