List your company

Mathematical Visual Question Answering on MathVision

73.3AccuracyGemini 2.5 Pro
8.258425.144242.0358.9158Mar 7, 2026Mar 12, 2026Mar 18, 2026Mar 23, 2026Mar 29, 2026Apr 3, 2026Apr 9, 2026

Evaluation results

MethodLinks
Gemini 2.5 Pro2026.0473.3
OpenVLThinkerV2Parameters=8B2026.0453.4
Qwen3-VL GRPORL Strategy=GRPO2026.0452.6
ARES-7BParameters=7B2026.0451.9
OVR-7BParameters=7B2026.0451.8
Qwen3-VL GDPORL Strategy=GDPO2026.0449.7
OneThinker-8BParameters=8B2026.0448.3
Qwen3-VL-Instruct-8BParameters=8B2026.0445.4
VL-Rethinker-7BParameters=7B2026.0432.3
Vision-G12026.0431.3
GPT-4o2026.0430.4
VisionZero2026.0428.5
MM-Eureka-7BParameters=7B2026.0426.9
VD-LF (HQ vs. LQ)Model=Qwen2.5-7B-VL, Train Set=VQA2026.0325.66
VD-LB (HQ Correct vs. LQ Wrong)Model=Qwen2.5-7B-VL, Train Set=VQA2026.0325.36
OpenVLThinker-7BParameters=7B2026.0425.3
VD-LF (HQ vs. LQ)Model=Qwen2.5-7B-VL, Train Set=GQA2026.0325.23
VD-LF (HQ vs. LQ)Model=Qwen2.5-7B-VL, Train Set=HiTab2026.0325.16
VD-LB (HQ Correct vs. LQ Wrong)Model=Qwen2.5-7B-VL, Train Set=HiTab2026.0325.13
BaselineModel=Qwen2.5-7B-VL, Training=Inference Only2026.0324.87
VD-LB (HQ Correct vs. LQ Wrong)Model=Qwen2.5-7B-VL, Train Set=GQA2026.0324.54
SFT (HQ Correct Only)Model=Qwen2.5-7B-VL, Train Set=HiTab2026.0322.76
VD-LF (HQ vs. LQ)Model=Qwen2.5-3B-VL, Train Set=HiTab2026.0318.78
VD-LB (HQ Correct vs. LQ Wrong)Model=Qwen2.5-3B-VL, Train Set=HiTab2026.0318.55
SFT (HQ Correct Only)Model=Qwen2.5-7B-VL, Train Set=VQA2026.0318.36
SFT (HQ Correct Only)Model=Qwen2.5-7B-VL, Train Set=GQA2026.0318.09
BaselineModel=Qwen2.5-3B-VL, Training=Inference Only2026.0317.76
VD-LB (HQ Correct vs. LQ Wrong)Model=Qwen2.5-3B-VL, Train Set=GQA2026.0317.5
VD-LF (HQ vs. LQ)Model=Qwen2.5-3B-VL, Train Set=VQA2026.0317.4
VD-LB (HQ Correct vs. LQ Wrong)Model=Qwen2.5-3B-VL, Train Set=VQA2026.0316.68
VD-LF (HQ vs. LQ)Model=Qwen2.5-3B-VL, Train Set=GQA2026.0316.68
SFT (HQ Correct Only)Model=Qwen2.5-3B-VL, Train Set=HiTab2026.0313.45
SFT (HQ Correct Only)Model=Qwen2.5-3B-VL, Train Set=VQA2026.0311.97
SFT (HQ Correct Only)Model=Qwen2.5-3B-VL, Train Set=GQA2026.0310.76