List your company
ResearchFormerly WizWand

Vision-centric Reasoning on MMStar-V

68.4AccuracyLong Grounded Thoughts (SFT + GRPO)
60.49662.54864.666.652Nov 7, 2025
Updated 8mo ago

Evaluation results

MethodLinks
Long Grounded Thoughts (SFT + GRPO)Open Model=Yes, Open Data=Yes, Base Model=Qwen2.5-VL-7B-Instruct2025.1168.4
MiMo-VL-7B-SFTOpen Model=Yes, Open Data=No2025.1167.6
MiMo-VL-7B-RLOpen Model=Yes, Open Data=No2025.1167.07
Long Grounded Thoughts (Multistage SFT + DPO)Open Model=Yes, Open Data=Yes, Base Model=Qwen2.5-VL-7B-Instruct2025.1166.27
Qwen2.5-VL-7B-InstructOpen Model=Yes, Open Data=No2025.1165.6
Long Grounded Thoughts (SFT + DPO)Open Model=Yes, Open Data=Yes, Base Model=Qwen2.5-VL-7B-Instruct2025.1164.4
Long Grounded Thoughts (SFT)Open Model=Yes, Open Data=Yes, Base Model=Qwen2.5-VL-7B-Instruct2025.1164.27
Qwen2.5-VL-7B-Instruct + LongPerceptualThoughtsOpen Model=Yes, Open Data=Yes2025.1164.13
Qwen2.5-VL-7B-Instruct + VLAA-ThinkerOpen Model=Yes, Open Data=Yes2025.1163.07
Qwen2.5-VL-7B-Instruct + Revisual-R1-finalOpen Model=Yes, Open Data=Yes2025.1160.8