| question-answering-on-boolq | MERIT (Proposed, 1D split, 2 groups, Qwen2.5-3B) | #18 | Accuracy: 86.3 |
| common-sense-reasoning-on-winogrande | MERIT (Proposed, 2D split, 4 groups, Qwen2.5-3B) | #74 | Accuracy: 70.8 |
| sentence-completion-on-hellaswag | MERIT (Proposed, 2D split, 4 groups, Qwen2.5-3B) | #94 | Accuracy: 74.7 |
| visual-question-answering-on-mm-vet | MERIT (Proposed, 2D split, Scaled base VLM 7B) | #105 | GPT-4 score: 39.1 |
| code-generation-on-humaneval | MERIT (Proposed, 2D split, 4 groups, Qwen2.5-3B) | #108 | Pass@1: 41.5 |
| multimodal-reasoning-on-mmmu | MERIT (Proposed, 3D split, 8 groups) | #118 | Accuracy: 42.7 |