List your company

General Visual Understanding on MMMU

75.1AccuracyClaude4-Sonnet
21.95635.75349.5563.347Aug 27, 2025Oct 9, 2025Nov 21, 2025Jan 3, 2026Feb 15, 2026Mar 30, 2026May 13, 2026

Evaluation results

MethodLinks
Claude4-Sonnet2026.0275.164.1
Qwen2.5-VL-72BParameters=72B2026.0270.261.9
DPEBackbone=Qwen3-VL-8B-Instruct, Iteration=32026.0269.1164.39
GPT-4o2026.0269.156.1
GPT5-Mini2026.0267.953.8
PDCRBackbone=Qwen3-VL-8B-Instruct2026.0557.1—
PACRBackbone=Qwen3-VL-8B-Instruct2026.0556.1—
DAPOBackbone=Qwen3-VL-8B-Instruct2026.0555.5—
Zero-shot InferenceBackbone=Qwen3-VL-8B-Instruct2026.0555.2—
GRPOBackbone=Qwen3-VL-8B-Instruct2026.0555—
Vision-SR1Backbone=Qwen2.5-VL-7B2025.0852.252.2
Vision-R1Backbone=Qwen2.5-VL-7B, Training Data=47K data2025.0851.850.7
PDCRBackbone=Qwen2.5-VL-7B2026.0551.5—
DAPOBackbone=Qwen2.5-VL-7B2026.0551—
GRPOBackbone=Qwen2.5-VL-7B2026.0550.7—
PACRBackbone=Qwen2.5-VL-7B2026.0550.5—
Vision-SR1Backbone=Qwen2.5-VL-3B2025.0849.648.8
Vision-R1Backbone=Qwen2.5-VL-3B, Training Data=47K data2025.0849.547.1
Vision-SR1Backbone=Mimo-VL-7B2025.0849.549.5
Vision-R1Backbone=Mimo-VL-7B, Training Data=47K data2025.0847.346
Zero-shot InferenceBackbone=Mimo-VL-7B, Inference Protocol=Zero-shot (before RL)2025.0845.644.4
PDCRBackbone=Qwen2.5-VL-3B2026.0543.2—
Vision-R1 (7B)Backbone=7B2025.0842.844.5
Vision-R1Backbone=Vision-R1 (7B)2026.0542.8—
PACRBackbone=Qwen2.5-VL-3B2026.0541.2—
DAPOBackbone=Qwen2.5-VL-3B2026.0541—
Percention-R1 (7B)Backbone=7B2025.0840.945.2
Perception-R1Backbone=Perception-R1 (7B)2026.0540.9—
GRPOBackbone=Qwen2.5-VL-3B2026.0540.4—
Zero-shot InferenceBackbone=Qwen2.5-VL-7B, Inference Protocol=Zero-shot (before RL)2025.0833.541.5
Zero-shot InferenceBackbone=Qwen2.5-VL-7B, Mode=Zero-shot2026.0533.3—
Visionary-R1 (3B)Backbone=3B2025.0830.633.9
Visionary-R1Backbone=Visionary-R1 (3B)2026.0530.6—
Zero-shot InferenceBackbone=Qwen2.5-VL-3B, Inference Protocol=Zero-shot (before RL)2025.0825.535.5
Zero-shot InferenceBackbone=Qwen2.5-VL-3B, Mode=Zero-shot2026.0524—
BaseTeacher=Qwen3-VL-8B, Student=Qwen3-VL-2B, Training Data=Geometry3K, Temperature=1.0, Checkpoint Selection=Best checkpoint2026.05—49.1
CoT-SFTTeacher=Qwen3-VL-8B, Student=Qwen3-VL-2B, Training Data=Geometry3K, Temperature=1.0, Checkpoint Selection=Best checkpoint2026.05—48.7
GRPOTeacher=Qwen3-VL-8B, Student=Qwen3-VL-2B, Training Data=Geometry3K, Temperature=1.0, Checkpoint Selection=Best checkpoint2026.05—52.9
Off-policy KDTeacher=Qwen3-VL-8B, Student=Qwen3-VL-2B, Training Data=Geometry3K, Temperature=1.0, Checkpoint Selection=Best checkpoint2026.05—49.3
OPDTeacher=Qwen3-VL-8B, Student=Qwen3-VL-2B, Training Data=Geometry3K, Temperature=1.0, Checkpoint Selection=Best checkpoint2026.05—50.9
PAPOTeacher=Qwen3-VL-8B, Student=Qwen3-VL-2B, Training Data=Geometry3K, Temperature=1.0, Checkpoint Selection=Best checkpoint2026.05—52.7
VA-OPDTeacher=Qwen3-VL-8B, Student=Qwen3-VL-2B, Training Data=Geometry3K, Temperature=1.0, Checkpoint Selection=Best checkpoint2026.05—51.5