| visual-question-answering-on-mmbench | LLaVA-InternLM2-ViT + MoSLoRA | #1 | GPT-3.5 score: 73.8 |
| visual-question-answering-on-mmbench | LLaVA-LLaMA3-8B-ViT + MoSLoRA | #3 | GPT-3.5 score: 73.0 |
| common-sense-reasoning-on-arc-easy | LLaMA 3 8B+MoSLoRA (fine-tuned) | #2 | Accuracy: 90.5 |
| question-answering-on-openbookqa | LLaMA-3 8B+MoSLoRA | #3 | Accuracy: 86.8 |
| question-answering-on-piqa | LLaMA3 8B+MoSLoRA | #3 | Accuracy: 89.7 |
| common-sense-reasoning-on-arc-challenge | LLaMA 3 8B + MoSLoRA (fine-tuned) | #5 | Accuracy: 81.5 |
| question-answering-on-social-iqa | LLaMA-3 8B+MoSLoRA (fine-tuned) | #5 | Accuracy: 81.0 |
| sentence-completion-on-hellaswag | LLaMA3+MoSLoRA | #5 | Accuracy: 95.0 |
| common-sense-reasoning-on-winogrande | LLaMA3 8B+MoSLoRA | #13 | Accuracy: 85.8 |
| question-answering-on-boolq | LLaMA3+MoSLoRA | #60 | Accuracy: 74.6 |
| visual-question-answering-on-mm-vet | LLaVA-InternLM2-7B-ViT + MoSLoRA | #139 | GPT-4 score: 35.2 |
| visual-question-answering-on-mm-vet | InternLM2+ViT (QMoSLoRA) | #140 | GPT-4 score: 35.2 |