List your company

Multimodal Math Reasoning on MathVista (avg@8 accuracy)

69.53Avg@8 AccuracyPAPO_G
52.94257.248561.55565.8615Jul 8, 2025

Evaluation results

MethodLinks
PAPO_GBackbone=Qwen2.5-VL, Model Size=7B, Training Algorithm=PAPO_G, Inference Temperature=1.02025.0769.53
PAPO_DBackbone=Qwen2.5-VL, Model Size=7B, Training Algorithm=PAPO_D, Inference Temperature=1.02025.0767.53
GRPOBackbone=Qwen2.5-VL, Model Size=7B, Training Algorithm=GRPO, Inference Temperature=1.02025.0765.48
PAPO_DBackbone=Qwen2.5-VL, Model Size=3B, Training Algorithm=PAPO_D, Inference Temperature=1.02025.0762.53
DAPOBackbone=Qwen2.5-VL, Model Size=7B, Training Algorithm=DAPO, Inference Temperature=1.02025.0761.91
PAPO_GBackbone=Qwen2.5-VL, Model Size=3B, Training Algorithm=PAPO_G, Inference Temperature=1.02025.0761.38
DAPOBackbone=Qwen2.5-VL, Model Size=3B, Training Algorithm=DAPO, Inference Temperature=1.02025.0760.89
GRPOBackbone=Qwen2.5-VL, Model Size=3B, Training Algorithm=GRPO, Inference Temperature=1.02025.0759.34
PAPO_GBackbone=Qwen3-VL (thinking), Model Size=2B, Training Algorithm=PAPO_G, Inference Temperature=1.02025.0756.08
GRPOBackbone=Qwen3-VL (thinking), Model Size=2B, Training Algorithm=GRPO, Inference Temperature=1.02025.0753.58