Visual Question Answering on LLaVA-W
49.1ROUGE-LAIM-CoT
Evaluation Results
| Method | Links | |
|---|---|---|
| AIM-CoTBackbone=Qwen2.5-VL-32B, Shot=1-shot2025.09 | 49.1 | |
| AIM-CoTBackbone=Qwen2.5-VL-32B, Shot=0-shot2025.09 | 46.5 | |
| ICoTBackbone=Qwen2.5-VL-32B, Shot=1-shot2025.09 | 44.7 | |
| ICoTBackbone=Qwen2.5-VL-32B, Shot=0-shot2025.09 | 43.4 | |
| CCoTBackbone=Qwen2.5-VL-32B, Shot=1-shot2025.09 | 42.4 | |
| No-CoTBackbone=Qwen2.5-VL-32B, Shot=1-shot2025.09 | 41.8 | |
| SCAFFOLDBackbone=Qwen2.5-VL-32B, Shot=1-shot2025.09 | 41.5 | |
| DDCoTBackbone=Qwen2.5-VL-32B, Shot=1-shot2025.09 | 41.3 | |
| No-CoTBackbone=Qwen2.5-VL-32B, Shot=0-shot2025.09 | 40.5 | |
| SCAFFOLDBackbone=Qwen2.5-VL-32B, Shot=0-shot2025.09 | 40.1 | |
| DDCoTBackbone=Qwen2.5-VL-32B, Shot=0-shot2025.09 | 39.6 | |
| MMCoTBackbone=Qwen2.5-VL-32B, Shot=1-shot2025.09 | 39.2 | |
| MMCoTBackbone=Qwen2.5-VL-32B, Shot=0-shot2025.09 | 38.9 | |
| CCoTBackbone=Qwen2.5-VL-32B, Shot=0-shot2025.09 | 37.8 | |
| AIM-CoTBackbone=Qwen2-VL-7B, Shot=1-shot2025.09 | 37.3 | |
| AIM-CoTBackbone=Janus-Pro-7B, Shot=1-shot2025.09 | 36.6 | |
| AIM-CoTBackbone=Qwen2-VL-7B, Shot=0-shot2025.09 | 36.3 | |
| ICoTBackbone=Qwen2-VL-7B, Shot=1-shot2025.09 | 35.7 | |
| AIM-CoTBackbone=Janus-Pro-7B, Shot=0-shot2025.09 | 35.5 | |
| ICoTBackbone=Qwen2-VL-7B, Shot=0-shot2025.09 | 34.2 | |
| CCoTBackbone=Qwen2-VL-7B, Shot=1-shot2025.09 | 33.9 | |
| ICoTBackbone=Janus-Pro-7B, Shot=1-shot2025.09 | 33.6 | |
| No-CoTBackbone=Qwen2-VL-7B, Shot=1-shot2025.09 | 33.5 | |
| SCAFFOLDBackbone=Qwen2-VL-7B, Shot=1-shot2025.09 | 33.1 | |
| DDCoTBackbone=Qwen2-VL-7B, Shot=1-shot2025.09 | 32.8 | |
| No-CoTBackbone=Qwen2-VL-7B, Shot=0-shot2025.09 | 32.7 | |
| ICoTBackbone=Janus-Pro-7B, Shot=0-shot2025.09 | 32.5 | |
| AIM-CoTBackbone=Chameleon-7B, Shot=1-shot2025.09 | 32 | |
| SCAFFOLDBackbone=Qwen2-VL-7B, Shot=0-shot2025.09 | 31.8 | |
| CCoTBackbone=Janus-Pro-7B, Shot=1-shot2025.09 | 31.5 | |
| MMCoTBackbone=Qwen2-VL-7B, Shot=1-shot2025.09 | 31.4 | |
| DDCoTBackbone=Qwen2-VL-7B, Shot=0-shot2025.09 | 31.2 | |
| No-CoTBackbone=Janus-Pro-7B, Shot=1-shot2025.09 | 30.9 | |
| SCAFFOLDBackbone=Janus-Pro-7B, Shot=1-shot2025.09 | 30.8 | |
| MMCoTBackbone=Qwen2-VL-7B, Shot=0-shot2025.09 | 30.7 | |
| DDCoTBackbone=Janus-Pro-7B, Shot=1-shot2025.09 | 30.4 | |
| AIM-CoTBackbone=Chameleon-7B, Shot=0-shot2025.09 | 29.8 | |
| No-CoTBackbone=Janus-Pro-7B, Shot=0-shot2025.09 | 29.8 | |
| SCAFFOLDBackbone=Janus-Pro-7B, Shot=0-shot2025.09 | 29.7 | |
| CCoTBackbone=Qwen2-VL-7B, Shot=0-shot2025.09 | 29.4 | |
| DDCoTBackbone=Janus-Pro-7B, Shot=0-shot2025.09 | 29 | |
| MMCoTBackbone=Janus-Pro-7B, Shot=1-shot2025.09 | 28.9 | |
| MMCoTBackbone=Janus-Pro-7B, Shot=0-shot2025.09 | 28.2 | |
| CCoTBackbone=Janus-Pro-7B, Shot=0-shot2025.09 | 27.8 | |
| ICoTBackbone=Chameleon-7B, Shot=1-shot2025.09 | 27.6 | |
| ICoTBackbone=Chameleon-7B, Shot=0-shot2025.09 | 25.2 | |
| SCAFFOLDBackbone=Chameleon-7B, Shot=1-shot2025.09 | 24.7 | |
| CCoTBackbone=Chameleon-7B, Shot=1-shot2025.09 | 24.5 | |
| No-CoTBackbone=Chameleon-7B, Shot=1-shot2025.09 | 23.9 | |
| DDCoTBackbone=Chameleon-7B, Shot=1-shot2025.09 | 23.1 | |
| CCoTBackbone=Chameleon-7B, Shot=0-shot2025.09 | 22.1 | |
| SCAFFOLDBackbone=Chameleon-7B, Shot=0-shot2025.09 | 21.7 | |
| MMCoTBackbone=Chameleon-7B, Shot=1-shot2025.09 | 20.6 | |
| MMCoTBackbone=Chameleon-7B, Shot=0-shot2025.09 | 20.4 | |
| DDCoTBackbone=Chameleon-7B, Shot=0-shot2025.09 | 20.2 | |
| No-CoTBackbone=Chameleon-7B, Shot=0-shot2025.09 | 13.1 |