Multimodal Reasoning on M^3CoT
58.7AccuracyDAP-ICoT
Evaluation Results
| Method | Links | |
|---|---|---|
| DAP-ICoTBackbone=Qwen2-VL-7B, Shots=1-Shot2026.03 | 58.7 | |
| DAP-ICoTBackbone=Qwen2-VL-7B, Shots=0-Shot2026.03 | 57.2 | |
| SCAFFOLDBackbone=Qwen2-VL-7B, Shots=1-Shot2026.03 | 53.6 | |
| CCoTBackbone=Qwen2-VL-7B, Shots=1-Shot2026.03 | 53 | |
| DAP-ICoTBackbone=Qwen2-VL-2B, Shots=1-Shot2026.03 | 51 | |
| SCAFFOLDBackbone=Qwen2-VL-7B, Shots=0-Shot2026.03 | 49.9 | |
| CCoTBackbone=Qwen2-VL-7B, Shots=0-Shot2026.03 | 48.7 | |
| MMCoTBackbone=Qwen2-VL-7B, Shots=1-Shot2026.03 | 47.5 | |
| DAP-ICoTBackbone=Qwen2-VL-2B, Shots=0-Shot2026.03 | 47.3 | |
| DDCoTBackbone=Qwen2-VL-7B, Shots=1-Shot2026.03 | 45.3 | |
| ICoTBackbone=Qwen2-VL-7B, Shots=1-Shot2026.03 | 44.8 | |
| MMCoTBackbone=Qwen2-VL-7B, Shots=0-Shot2026.03 | 44.4 | |
| DDCoTBackbone=Qwen2-VL-7B, Shots=0-Shot2026.03 | 43.9 | |
| SCAFFOLDBackbone=Qwen2-VL-2B, Shots=1-Shot2026.03 | 43.6 | |
| DAP-ICoTBackbone=Chameleon-7B, Shots=1-Shot2026.03 | 41.9 | |
| DAP-ICoTBackbone=LLaVA-V1.5-13B, Shots=1-Shot2026.03 | 41.8 | |
| DAP-ICoTBackbone=Chameleon-7B, Shots=0-Shot2026.03 | 41 | |
| SCAFFOLDBackbone=Qwen2-VL-2B, Shots=0-Shot2026.03 | 40.3 | |
| DAP-ICoTBackbone=LLaVA-V1.5-13B, Shots=0-Shot2026.03 | 39.4 | |
| DDCoTBackbone=Qwen2-VL-2B, Shots=1-Shot2026.03 | 39.3 | |
| ICoTBackbone=Qwen2-VL-7B, Shots=0-Shot2026.03 | 38 | |
| ICoTBackbone=LLaVA-V1.5-13B, Shots=1-Shot2026.03 | 37.9 | |
| DDCoTBackbone=Qwen2-VL-2B, Shots=0-Shot2026.03 | 37.9 | |
| CCoTBackbone=Qwen2-VL-2B, Shots=1-Shot2026.03 | 37.7 | |
| DAP-ICoTBackbone=LLaVA-V1.5-7B, Shots=1-Shot2026.03 | 37.6 | |
| ICoTBackbone=Qwen2-VL-2B, Shots=1-Shot2026.03 | 37.3 | |
| ICoTBackbone=LLaVA-V1.5-13B, Shots=0-Shot2026.03 | 37 | |
| DAP-ICoTBackbone=LLaVA-V1.5-7B, Shots=0-Shot2026.03 | 36.3 | |
| ICoTBackbone=Qwen2-VL-2B, Shots=0-Shot2026.03 | 35.8 | |
| CCoTBackbone=LLaVA-V1.5-7B, Shots=1-Shot2026.03 | 35.5 | |
| ICoTBackbone=LLaVA-V1.5-7B, Shots=1-Shot2026.03 | 35 | |
| DirectBackbone=Qwen2-VL-7B, Shots=1-Shot2026.03 | 35 | |
| ICoTBackbone=LLaVA-V1.5-7B, Shots=0-Shot2026.03 | 34.6 | |
| CCoTBackbone=LLaVA-V1.5-7B, Shots=0-Shot2026.03 | 34.2 | |
| DDCoTBackbone=LLaVA-V1.5-13B, Shots=1-Shot2026.03 | 33.8 | |
| SCAFFOLDBackbone=LLaVA-V1.5-13B, Shots=1-Shot2026.03 | 33.2 | |
| DirectBackbone=Qwen2-VL-7B, Shots=0-Shot2026.03 | 33 | |
| MMCoTBackbone=LLaVA-V1.5-13B, Shots=1-Shot2026.03 | 32.9 | |
| DDCoTBackbone=LLaVA-V1.5-13B, Shots=0-Shot2026.03 | 32.9 | |
| ICoTBackbone=Chameleon-7B, Shots=1-Shot2026.03 | 32.1 | |
| MMCoTBackbone=LLaVA-V1.5-13B, Shots=0-Shot2026.03 | 32.1 | |
| CCoTBackbone=LLaVA-V1.5-13B, Shots=1-Shot2026.03 | 32 | |
| SCAFFOLDBackbone=LLaVA-V1.5-13B, Shots=0-Shot2026.03 | 31.9 | |
| SCAFFOLDBackbone=Chameleon-7B, Shots=1-Shot2026.03 | 31.2 | |
| SCAFFOLDBackbone=Chameleon-7B, Shots=0-Shot2026.03 | 31 | |
| MMCoTBackbone=LLaVA-V1.5-7B, Shots=1-Shot2026.03 | 30.4 | |
| DDCoTBackbone=Chameleon-7B, Shots=1-Shot2026.03 | 30.3 | |
| CCoTBackbone=LLaVA-V1.5-13B, Shots=0-Shot2026.03 | 30.1 | |
| DDCoTBackbone=Chameleon-7B, Shots=0-Shot2026.03 | 29.8 | |
| MMCoTBackbone=LLaVA-V1.5-7B, Shots=0-Shot2026.03 | 29.6 | |
| SCAFFOLDBackbone=LLaVA-V1.5-7B, Shots=1-Shot2026.03 | 28.3 | |
| MMCoTBackbone=Chameleon-7B, Shots=1-Shot2026.03 | 28 | |
| SCAFFOLDBackbone=LLaVA-V1.5-7B, Shots=0-Shot2026.03 | 26.6 | |
| DDCoTBackbone=LLaVA-V1.5-7B, Shots=1-Shot2026.03 | 26.5 | |
| CCoTBackbone=Chameleon-7B, Shots=1-Shot2026.03 | 26.3 | |
| ICoTBackbone=Chameleon-7B, Shots=0-Shot2026.03 | 26.1 | |
| MMCoTBackbone=Chameleon-7B, Shots=0-Shot2026.03 | 26 | |
| DirectBackbone=LLaVA-V1.5-7B, Shots=1-Shot2026.03 | 25.7 | |
| DirectBackbone=LLaVA-V1.5-13B, Shots=1-Shot2026.03 | 25.5 | |
| DDCoTBackbone=LLaVA-V1.5-7B, Shots=0-Shot2026.03 | 25.4 | |
| CCoTBackbone=Chameleon-7B, Shots=0-Shot2026.03 | 25.1 | |
| DirectBackbone=LLaVA-V1.5-13B, Shots=0-Shot2026.03 | 24.6 | |
| DirectBackbone=Qwen2-VL-2B, Shots=1-Shot2026.03 | 24.5 | |
| DirectBackbone=Chameleon-7B, Shots=1-Shot2026.03 | 23.8 | |
| DirectBackbone=LLaVA-V1.5-7B, Shots=0-Shot2026.03 | 23.2 | |
| DirectBackbone=Chameleon-7B, Shots=0-Shot2026.03 | 22.5 | |
| MMCoTBackbone=Qwen2-VL-2B, Shots=1-Shot2026.03 | 22.4 | |
| CCoTBackbone=Qwen2-VL-2B, Shots=0-Shot2026.03 | 20.2 | |
| MMCoTBackbone=Qwen2-VL-2B, Shots=0-Shot2026.03 | 14.9 | |
| DirectBackbone=Qwen2-VL-2B, Shots=0-Shot2026.03 | 14.4 |