Multi-Image Understanding on MuirBench (test)

68Accuracy

GPT-4o

Updated 1mo ago

Evaluation Results

Method	Links
GPT-4o 2024.12		68
GPT-4V 2024.12		62.3
GPT-4V 2026.01		62.3
InternVL-2-8B 2024.12		59.4
MAmmoTH-VL-8B 2024.12		55.1
LLaVA-OV-72B 2024.12		54.8
attention-masking strategy 2026.01		51.3
InternVL2-Llama3-76B 2026.01		51.2
Qwen2VL-7B 2026.01		43
LLaVA-OV-7B 2024.12		41.8
LLaVA-OV-7B 2026.01		41.7
Qwen2-VL-7B-Ins. 2024.12		41.6
InternVL2-8B 2026.01		37.9
procedural data-generation strategy 2026.01		33.6
LLaVA-OV-72B 2024.12		33.2
LLaVA-OV-7B 2024.12		32.7
attention-masking strategy 2026.01		32.5
Qwen2VL-2B 2026.01		27.2
LLaVA-OV-0.5B 2026.01		26.8
InternVL2-2B 2026.01		24.3
LLaVA-v1.5-7B 2026.01		20