List your company

Perception-intensive Reasoning on MME-RealWorld-Lite

55.13ScoreStarve to Perceive
26.249233.747141.24548.7429May 18, 2026

Evaluation results

MethodLinks
Starve to PerceiveVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.0555.13
Qwen2.5-VL + BA-SFT + VanillaRLVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.0553.78
DeepEyesVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.0550.81
ChainOfFocusVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.0548.98
Starve to PerceiveVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.0547.37
PixelReasonerVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.0547.21
Qwen2.5-VL + BA-SFTVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.0546.95
Qwen2.5-VL + BA-SFT + VanillaRLVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.0546.69
GPT-4oVisual Bandwidth Constraint (B)=No Constrain, Reported From=Other publications2026.0546.4
LLaVA-OneVisionVisual Bandwidth Constraint (B)=No Constrain, Reported From=Other publications2026.0543.7
Qwen2.5-VLVisual Bandwidth Constraint (B)=No Constrain, Reported From=Other publications2026.0543
DeepEyesVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.0540.85
Qwen2.5-VL + BA-SFTVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.0539.55
ChainOfFocusVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.0537.83
Mini-o3Visual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.0532.15
PixelReasonerVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.0529.6
Qwen2.5-VLVisual Bandwidth Constraint (B)=2562026.0527.41
Mini-o3Visual Bandwidth Constraint (B)=256, Tool-calling=True2026.0527.36