List your company

Math Reasoning on We-Math

85.6Avg Pass@8ADHint
34.84848.02461.274.376Dec 15, 2025Jan 10, 2026Feb 5, 2026Mar 3, 2026Mar 29, 2026Apr 24, 2026May 21, 2026

Evaluation results

MethodLinks
ADHintBase Model=Qwen3-VL-8B-Instruct2025.1285.676.4
GRPOBase Model=Qwen3-VL-8B-Instruct2025.1282.673.1
SFT -> RLBase MLLM=Qwen2.5-VL-7B2025.1272.554.1
HintGRPOBase Model=Qwen3-VL-8B-Instruct2025.1271.768.6
ADHintBase MLLM=Qwen2.5-VL-7B2025.1270.869.9
SFTBase MLLM=Qwen2.5-VL-7B2025.1270.144.1
GRPOBase MLLM=Qwen2.5-VL-7B2025.1268.869.8
SFT -> RLBase MLLM=Qwen2.5-VL-3B2025.1260.638.3
ADHintBase MLLM=Qwen2.5-VL-3B2025.1259.160.4
GRPOBase MLLM=Qwen2.5-VL-3B2025.125959
SFTBase MLLM=Qwen2.5-VL-3B2025.1258.532.5
LUFFYBase MLLM=Qwen2.5-VL-7B2025.1254.852.2
GHPOBase MLLM=Qwen2.5-VL-7B2025.1252.863
VA-OPDTeacher=Qwen3-VL-8B, Student=Qwen3-VL-2B, Training Data=Geometry3K, Temperature=1.0, Checkpoint Selection=Best checkpoint2026.0546.6—
LUFFYBase MLLM=Qwen2.5-VL-3B2025.1246.142.3
PAPOTeacher=Qwen3-VL-8B, Student=Qwen3-VL-2B, Training Data=Geometry3K, Temperature=1.0, Checkpoint Selection=Best checkpoint2026.0544.8—
HintGRPOBase MLLM=Qwen2.5-VL-7B2025.1244.742.9
GRPOTeacher=Qwen3-VL-8B, Student=Qwen3-VL-2B, Training Data=Geometry3K, Temperature=1.0, Checkpoint Selection=Best checkpoint2026.0544.1—
OPDTeacher=Qwen3-VL-8B, Student=Qwen3-VL-2B, Training Data=Geometry3K, Temperature=1.0, Checkpoint Selection=Best checkpoint2026.0543.3—
StepHintBase MLLM=Qwen2.5-VL-7B2025.1241.139.8
HintGRPOBase MLLM=Qwen2.5-VL-3B2025.1240.542
StepHintBase MLLM=Qwen2.5-VL-3B2025.1240.232
CoT-SFTTeacher=Qwen3-VL-8B, Student=Qwen3-VL-2B, Training Data=Geometry3K, Temperature=1.0, Checkpoint Selection=Best checkpoint2026.0539.7—
GHPOBase MLLM=Qwen2.5-VL-3B2025.1239.141
Off-policy KDTeacher=Qwen3-VL-8B, Student=Qwen3-VL-2B, Training Data=Geometry3K, Temperature=1.0, Checkpoint Selection=Best checkpoint2026.0538.9—
BaseTeacher=Qwen3-VL-8B, Student=Qwen3-VL-2B, Training Data=Geometry3K, Temperature=1.0, Checkpoint Selection=Best checkpoint2026.0536.8—