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

Visual Mathematical Reasoning on WeMath strict

44.8Accuracy

AdaVaR-7B

Updated 19d ago

Evaluation Results

Method	Links
AdaVaR-7B 2025.09		44.8	55.82
OVR-7B 2025.09		44.6	55.76
GPT-4o 2025.09		42.9	53.2
VLAA-Thinker-7B 2025.09		41.5	54.2
InternVL3-8B 2025.09		37.5	-
MM-Eureka 2025.09		36.9	52.52
ViGoRL-7B 2025.09		36.3	50.53
VLAA-Thinker-3B 2025.09		33.8	46.7
AdaVaR-3B 2025.09		33.8	50.84
Chain-of-Focus 2025.09		33.1	50.94
DeepEyes 2025.09		32.7	53.72
Orsta-7B 2025.09		31.8	52.04
Qwen2.5-VL-7B 2025.09		31.2	50.9
LMM-R1 2025.09		29.2	48.8
GRIT-3B 2025.09		29.1	45.8
ViGoRL-3B 2025.09		28.5	45.7
Qwen2.5-VL-3B 2025.09		21.2	45.74
LLaVA-OV-7B 2025.09		17.7	-