| Benchmark | Model | Rank | Results |
|---|---|---|---|
| code-generation-on-humaneval | LoopCD-Ouro-2.6B (adaptive) | – | Pass@1: 79.88 |
| common-sense-reasoning-on-winogrande | LoopCD-Ouro-2.6B (adaptive) | – | Accuracy: 72.31 |
| multi-task-language-understanding-on-mmlu-pro | LoopCD-Ouro-2.6B (fixed) | – | Accuracy: 57.5 |
| question-answering-on-piqa | LoopCD-Ouro-2.6B (adaptive) | – | Accuracy: 80.69 |
| sentence-completion-on-hellaswag | LoopCD-Ouro-2.6B (adaptive) | – | Accuracy: 81.66 |