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

Multimodal Math Reasoning on R1-OV Bench

38.9Accuracy

AVATAR

Updated 19d ago

Evaluation Results

Method	Links
AVATAR 2025.08		38.9
AVATAR 2025.08		38.5
VLAAThinker-VL-7B 2025.08		38.4
AVATAR 2025.08		37.5
AVATAR 2025.08		37
Qwen2.5-VL-7B + GRPO 2025.08		36.8
Qwen2.5-VL-7B 2025.08		34.9
OpenVLThinker-7B 2025.08		34.7