List your company

Medical Visual Question Answering on Medical VQA

96.76Accuracy (CT)Evo-PI
15.567236.646157.72578.8039Jun 30, 2026

Evaluation results

MethodLinks
Evo-PIBase Model=Med-R1 [12], RLVR Algorithm=GRPO2026.0696.7693.1991.9587.294.6298.999.8291.0194.12
Evo-PIBase Model=Med-R1 [12], RLVR Algorithm=GSPO2026.0696.2893.3490.8187.3992.9899.3499.8290.3893.91
Evo-PIBase Model=HuatuoGPT-Vision [11], RLVR Algorithm=GSPO2026.0687.9790.2193.6995.891.3898.2491.0993.6792.95
Evo-PIBase Model=HuatuoGPT-Vision [11], RLVR Algorithm=GRPO2026.0687.9790.2193.6995.894.1898.2491.0993.6792.95
Med-R1Base Model=Qwen 2-VL-2B, Modality-specific submodels=true2026.0671.683.3890.1974.4751.4489.4677.7378.5477.1
Qwen 2.5-VL-3BBase Model=Qwen 2.5-VL-3B2026.0671.0360.7869.8159.8653.6468.3944.0974.6262.78
Qwen 2-VL-7BBase Model=Qwen 2-VL-7B2026.0668.1857.1974.4462.554.5263.2338.2771.9761.29
Qwen 2-VL-72BBase Model=Qwen 2-VL-72B2026.0667.9765.3172.5867.8469.3972.7651.3972.1168.05
Qwen 2.5-VL-7BBase Model=Qwen 2.5-VL-7B2026.0667.3670.0670.5660.0455.3355.733.5575.6361.03
Qwen 2.5-VL-72BBase Model=Qwen 2.5-VL-72B2026.0666.1869.7571.0469.3763.6469.2269.8579.8167.71
HuatuoGPT-VisionBase Model=Qwen 2.5-VL-7B2026.0665.3468.4176.371.368.6677.6348.1880.0569.48
MedVInT2026.0640.7429.1131.8432.0243.123.2641.2655.137.05
Qwen 2-VL-2BBase Model=Qwen 2-VL-2B2026.0640.2341.7746.8542.0844.138.2838.6448.2342.52
Med-Flamingo2026.0631.2848.5641.2630.0326.3425.1631.6944.0134.29
RadFM2026.0627.5639.2136.8627.9724.0632.816.5730.9529.5
LLaVA-Med2026.0618.6944.9539.0333.2927.4734.6129.8830.6832.33