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

Question Answering on WTQ (held-in)

80.34Accuracy

Gemini-2.5-Pro

Updated 22d ago

Evaluation Results

Method	Links
Gemini-2.5-Pro 2025.09		80.34
OpenAI-o4-mini 2025.09		78.2
Claude-3.5-Sonnet 2025.09		71.8
QVQ-72B-Preview 2025.09		68.2
Qwen2-VL-72B 2025.09		65.7
GPT-4o 2025.09		64.39
Qwen2-VL-7B-GRPO 2025.09		53.2
Qwen2-VL-7B-Table-R1 2025.09		50.3
Qwen2-VL-7B-SFT 2025.09		46.8
Qwen2-VL-7B 2025.09		46.5
DeepSeek-VL2 4.5B 2025.09		40.42
Qwen2-VL-2B-GRPO 2025.09		35.3
Qwen2-VL-2B-Table-R1 2025.09		34.4
Qwen2-VL-2B-SFT 2025.09		31
Qwen2-VL-2B 2025.09		22.3
Table-LLaVA 13B 2025.09		20.41
Qwen2-VL-7B-Ins 2025.09		19.73
Monkey 7B 2025.09		19.07
Table-LLaVA 7B 2025.09		18.43
LLaVA-v1.5 7B 2025.09		1.24
mPLUG-Owl2 7B 2025.09		0.67