List your company

Mathematical Reasoning on MathVista (ASR, Coh, Rat, Acc)

77.9AccuracyQwen3-VL-4B-Instruct / H-OPD (8B)
43.68452.56761.4570.333Jul 1, 2026

Evaluation results

MethodLinks
Qwen3-VL-4B-Instruct / H-OPD (8B)Student=Qwen3-VL-4B-Instruct, Teacher=Qwen3-VL-8B + Qwen3-8B, Method=H-OPD2026.0777.9————
Qwen3-VL-8BType=Baseline2026.0777.2————
Qwen3-VL-4B-Instruct / ExOPD (8B)Student=Qwen3-VL-4B-Instruct, Teacher=Qwen3-VL-8B, Method=ExOPD2026.0777.2————
Qwen3-VL-4B-Instruct / OPD (8B)Student=Qwen3-VL-4B-Instruct, Teacher=Qwen3-VL-8B, Method=OPD2026.0776.6————
Qwen3-VL-4B-Instruct / GRPOStudent=Qwen3-VL-4B-Instruct, Teacher=Self, Method=GRPO2026.0776.1————
Qwen3-VL-4BType=Baseline2026.0773.7————
Qwen3-VL-4B-Instruct / SFT (8B)Student=Qwen3-VL-4B-Instruct, Teacher=Qwen3-VL-8B, Method=SFT2026.0772.1————
Qwen3-VL-2B-Instruct / H-OPD (8B)Student=Qwen3-VL-2B-Instruct, Teacher=Qwen3-VL-8B + Qwen3-8B, Method=H-OPD2026.0767.2————
Qwen3-VL-2B-Instruct / H-OPD (4B)Student=Qwen3-VL-2B-Instruct, Teacher=Qwen3-VL-4B + Qwen3-4B, Method=H-OPD2026.0766.8————
Qwen3-VL-2B-Instruct / ExOPD (8B)Student=Qwen3-VL-2B-Instruct, Teacher=Qwen3-VL-8B, Method=ExOPD2026.0766.1————
Qwen3-VL-2B-Instruct / OPD (8B)Student=Qwen3-VL-2B-Instruct, Teacher=Qwen3-VL-8B, Method=OPD2026.0765.7————
Qwen3-VL-2B-Instruct / ExOPD (4B)Student=Qwen3-VL-2B-Instruct, Teacher=Qwen3-VL-4B, Method=ExOPD2026.0764.4————
Qwen3-VL-2B-Instruct / OPD (4B)Student=Qwen3-VL-2B-Instruct, Teacher=Qwen3-VL-4B, Method=OPD2026.0764.2————
Qwen3-VL-2B-Instruct / GRPOStudent=Qwen3-VL-2B-Instruct, Teacher=Self, Method=GRPO2026.0763.6————
Qwen3-VL-2B-Instruct / SFT (8B)Student=Qwen3-VL-2B-Instruct, Teacher=Qwen3-VL-8B, Method=SFT2026.0762.2————
Qwen3-VL-2B-Instruct / SFT (4B)Student=Qwen3-VL-2B-Instruct, Teacher=Qwen3-VL-4B, Method=SFT2026.0761.6————
Qwen3-VL-2BType=Baseline2026.0761.3————
BaseBase model=Qwen2.5-VL-7B2026.0749—3.423.19—
ReShiftBase model=Qwen2.5-VL-7B2026.0749853.022.997
RewriteBase model=Qwen2.5-VL-7B2026.0747742.682.9224
BadVisionBase model=Qwen2.5-VL-7B2026.0746302.672.7913
BadTokenBase model=Qwen2.5-VL-7B2026.0745772.862.7222