Share your thoughts, 1 month free Claude Pro on usSee more

Fact Verification on InfoTabs (held-out)

79.5Accuracy

GPT-4o

Updated 22d ago

Evaluation Results

Method	Links
GPT-4o 2025.09		79.5
OpenAI-o4-mini 2025.09		78.3
Gemini-2.5-Pro 2025.09		77.15
QVQ-72B-Preview 2025.09		74.64
Qwen2-VL-72B 2025.09		72.82
Claude-3.5-Sonnet 2025.09		70.3
Table-LLaVA 13B 2025.09		66.91
Table-LLaVA 7B 2025.09		65.26
Qwen2-VL-7B-Table-R1 2025.09		62.8
Qwen2-VL-7B-SFT 2025.09		57.6
Qwen2-VL-7B-Ins 2025.09		46.56
Qwen2-VL-2B-Table-R1 2025.09		43.6
Qwen2-VL-7B-GRPO 2025.09		41.6
Qwen2-VL-2B-SFT 2025.09		37.4
Qwen2-VL-7B 2025.09		32.8
Qwen2-VL-2B-GRPO 2025.09		29
LLaVA-v1.5 7B 2025.09		28.31
mPLUG-Owl2 7B 2025.09		26.19
DeepSeek-VL2 4.5B 2025.09		24.79
Qwen2-VL-2B 2025.09		24.6
Monkey 7B 2025.09		22.11