Knowledge-based Vision-Language Understanding on Knowledge
48.1Average ScoreOurs w/o Masking
Evaluation Results
| Method | Links | |
|---|---|---|
| Ours w/o MaskingLLM=Qwen2-7B2025.12 | 48.1 | |
| JARVISLLM=Qwen2-7B2025.12 | 47.7 | |
| VIRALLLM=Qwen2-7B2025.12 | 47.2 | |
| LLaVALLM=Qwen2-7B2025.12 | 46.7 | |
| VIRALLLM=Vicuna-7B2025.12 | 42.6 | |
| JARVISLLM=Vicuna-7B2025.12 | 41.6 | |
| LLaVALLM=Vicuna-7B2025.12 | 41.4 | |
| Ours w/o MaskingLLM=Vicuna-7B2025.12 | 41.3 |