Correction on VISCO full 1.0 (test)
78Correction GainGemini-1.5-Pro
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini-1.5-ProSize=Unknown, Critique Source=Human2024.12 | 78 | |
| GPT-4oSize=Unknown, Critique Source=Human2024.12 | 76.2 | |
| Claude-3.5-SonnetSize=Unknown, Critique Source=Human2024.12 | 73.7 | |
| InternVL2Size=76B, Critique Source=Human2024.12 | 72.7 | |
| Llama-3.2Size=90B, Critique Source=Human2024.12 | 66.4 | |
| InternVL2Size=26B, Critique Source=Human2024.12 | 59.3 | |
| LLaVA-CriticSize=72B, Critique Source=Human2024.12 | 58.9 | |
| MolmoSize=72B, Critique Source=Human2024.12 | 53.1 | |
| MiniCPM-V2.6Size=8B, Critique Source=Human2024.12 | 53 | |
| InternVL2Size=8B, Critique Source=Human2024.12 | 52.7 | |
| Qwen2-VLSize=7B, Critique Source=Human2024.12 | 50.8 | |
| MolmoSize=7B, Critique Source=Human2024.12 | 49.1 | |
| InternVL2Size=40B, Critique Source=Human2024.12 | 47.6 | |
| NVLMSize=72B, Critique Source=Human2024.12 | 42.2 | |
| LLaVA-v1.6Size=7B, Critique Source=Human2024.12 | 40.3 | |
| LLaVA-v1.6Size=13B, Critique Source=Human2024.12 | 40.2 | |
| LLaVA-v1.6Size=34B, Critique Source=Human2024.12 | 39.2 | |
| Llama-3.2Size=11B, Critique Source=Human2024.12 | 34.8 | |
| LLaVA-OVSize=72B, Critique Source=Human2024.12 | 33.4 | |
| Qwen2-VLSize=72B, Critique Source=Human2024.12 | 31.5 | |
| GPT-4oSize=Unknown, Critique Source=Self2024.12 | 28.8 | |
| Claude-3.5-SonnetSize=Unknown, Critique Source=Self2024.12 | 25.6 | |
| Gemini-1.5-ProSize=Unknown, Critique Source=Self2024.12 | 24.9 | |
| LLaVA-OVSize=7B, Critique Source=Human2024.12 | 22.4 | |
| LLaVA-CriticSize=7B, Critique Source=Human2024.12 | 19.3 | |
| LLaVA-CriticSize=72B, Critique Source=Self2024.12 | 15.4 | |
| InternVL2Size=76B, Critique Source=Self2024.12 | 11.7 | |
| InternVL2Size=40B, Critique Source=Self2024.12 | 9.9 | |
| MiniCPM-V2.6Size=8B, Critique Source=Self2024.12 | 7.8 | |
| InternVL2Size=26B, Critique Source=Self2024.12 | 6 | |
| Qwen2-VLSize=7B, Critique Source=Self2024.12 | 5.5 | |
| InternVL2Size=8B, Critique Source=Self2024.12 | 5.4 | |
| Llama-3.2Size=90B, Critique Source=Self2024.12 | 4.4 | |
| DeepSeek-VLSize=7B, Critique Source=Human2024.12 | 2.3 | |
| MolmoSize=7B, Critique Source=Self2024.12 | 1.8 | |
| NVLMSize=72B, Critique Source=Self2024.12 | 1.7 | |
| MolmoSize=72B, Critique Source=Self2024.12 | 1.4 | |
| LLaVA-v1.6Size=34B, Critique Source=Self2024.12 | -1.7 | |
| Qwen2-VLSize=72B, Critique Source=Self2024.12 | -2.1 | |
| LLaVA-v1.6Size=13B, Critique Source=Self2024.12 | -7.2 | |
| LLaVA-v1.6Size=7B, Critique Source=Self2024.12 | -8.7 | |
| LLaVA-OVSize=7B, Critique Source=Self2024.12 | -9.1 | |
| LLaVA-OVSize=72B, Critique Source=Self2024.12 | -10.2 | |
| LLaVA-CriticSize=7B, Critique Source=Self2024.12 | -11.4 | |
| Llama-3.2Size=11B, Critique Source=Self2024.12 | -11.7 | |
| DeepSeek-VLSize=7B, Critique Source=Self2024.12 | -17.9 |