List your company
ResearchFormerly WizWand

Comprehensive Reasoning on Aggregate Benchmarks

66.9Average AccuracyQwen2.5-VL-7B + DUPL
51.92455.81259.763.588Oct 1, 2025
Updated 4mo ago

Evaluation results

MethodLinks
Qwen2.5-VL-7B + DUPLBase Model=Qwen2.5-VL-7B, RL Algorithm=DUPL2025.1066.9
Qwen2.5-VL-7B + NoisyRolloutBase Model=Qwen2.5-VL-7B, RL Algorithm=NoisyRollout2025.1065.4
PAPO-7BBase Model=PAPO-7B2025.1065.2
Qwen2.5-VL-7B + GRPOBase Model=Qwen2.5-VL-7B, RL Algorithm=GRPO2025.1063.9
VLAA-Thinker-7BBase Model=VLAA-Thinker-7B2025.1063.6
MM-Eureka-Qwen-7BBase Model=MM-Eureka-Qwen-7B2025.1063.5
OpenVLThinker-7BBase Model=OpenVLThinker-7B2025.1061.9
Qwen2.5-VL-3B + DUPLBase Model=Qwen2.5-VL-3B, RL Algorithm=DUPL2025.1061.2
Qwen2.5-VL-7BBase Model=Qwen2.5-VL-7B2025.1061.1
R1-Onevision-7BBase Model=R1-Onevision-7B2025.1060.5
Qwen2.5-VL-3B + NoisyRolloutBase Model=Qwen2.5-VL-3B, RL Algorithm=NoisyRollout2025.1059.6
Qwen2.5-VL-3B + GRPOBase Model=Qwen2.5-VL-3B, RL Algorithm=GRPO2025.1058.4
Qwen2.5-VL-3BBase Model=Qwen2.5-VL-3B2025.1052.5