| object-hallucination-evaluation-on-pope | LLaVA-NeXT-34B | #9 | F1: 87.73 |
| visual-question-answering-vqa-on-mme | LLaVA-NeXT-34B | #13 | Score: 2028 |
| visual-question-answering-vqa-on-seed-seedbench | LLaVA-NeXT-34B | #13 | Accuracy: 75.9 |
| visual-question-answering-vqa-on-mmbench-cn | LLaVA-NeXT-34B | #14 | Accuracy: 79 |
| visual-question-answering-vqa-on-mmbench-en | LLaVA-NeXT-34B | #23 | Accuracy: 79.3 |
| visual-question-answering-on-mm-vet | LLaVA-NeXT-34B | #52 | GPT-4 score: 57.4 |
| multimodal-reasoning-on-mmmu | LLaVA-NeXT-34B | #95 | Accuracy: 51.1 |
| multimodal-reasoning-on-mathvista | LLaVA-NeXT-34B | #108 | Accuracy: 46.5 |