Multi-image Reasoning on Mantis-Eval
78.3Overall ScoreInternVL2.5-38B
Evaluation Results
| Method | Links | |
|---|---|---|
| InternVL2.5-38BModel Name=InternVL2.5-38B2024.12 | 78.3 | |
| LLaVA-OneVision-72BModel Name=LLaVA-OneVision-72B2024.12 | 77.6 | |
| InternVL2.5-78BModel Name=InternVL2.5-78B2024.12 | 77 | |
| InternVL2.5-26BModel Name=InternVL2.5-26B2024.12 | 75.6 | |
| InternVL2-Llama3-76BModel Name=InternVL2-Llama3-76B2024.12 | 73.7 | |
| InternVL2-40BModel Name=InternVL2-40B2024.12 | 71.4 | |
| InternVL2-26BModel Name=InternVL2-26B2024.12 | 69.6 | |
| MiniCPM-V2.6Model Name=MiniCPM-V2.62024.12 | 69 | |
| InternVL2.5-8BModel Name=InternVL2.5-8B2024.12 | 67.7 | |
| InternVL-Chat-V1.5Model Name=InternVL-Chat-V1.52024.12 | 66.8 | |
| InternVL2-8BModel Name=InternVL2-8B2024.12 | 65.4 | |
| mPLUG-Owl3# Param=8B2024.08 | 63.1 | |
| LLAVA-Interleave# Param=8B2024.08 | 62.7 | |
| InternVL2.5-4BModel Name=InternVL2.5-4B2024.12 | 62.7 | |
| GPT-4VModel Name=GPT-4V2024.12 | 62.7 | |
| InternVL2-4BModel Name=InternVL2-4B2024.12 | 61.3 | |
| Mantis-SigLIP# Param=8B2024.08 | 59.5 | |
| InternVL2.5-2BModel Name=InternVL2.5-2B2024.12 | 54.8 | |
| VILA# Param=8B2024.08 | 51.2 | |
| InternVL2.5-1BModel Name=InternVL2.5-1B2024.12 | 51.2 | |
| Idefics2# Param=8B2024.08 | 48.9 | |
| InternVL2-2BModel Name=InternVL2-2B2024.12 | 48.4 | |
| InternVL2-1BModel Name=InternVL2-1B2024.12 | 46.1 | |
| InstructBLIP# Param=8B2024.08 | 45.6 | |
| CogVLM# Param=17B2024.08 | 45.2 | |
| LLaVA-OneVision-0.5BModel Name=LLaVA-OneVision-0.5B2024.12 | 39.6 | |
| Qwen-VL-Chat# Param=8B2024.08 | 39.2 | |
| VideoLLaVA# Param=8B2024.08 | 35.9 |