Visual Search and Perception-intensive Reasoning on Multi-benchmark Score Overall
64.59Overall ScoreStarve to Perceive
Evaluation Results
| Method | Links | |
|---|---|---|
| Starve to PerceiveVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.05 | 64.59 | |
| Qwen2.5-VL + BA-SFT + VanillaRLVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.05 | 64.48 | |
| DeepEyesVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.05 | 61.77 | |
| ChainOfFocusVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.05 | 61.75 | |
| PixelReasonerVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.05 | 60.1 | |
| Qwen2.5-VL + BA-SFTVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.05 | 58.95 | |
| Qwen2.5-VLVisual Bandwidth Constraint (B)=No Constrain, Reported From=Other publications2026.05 | 52.53 | |
| Starve to PerceiveVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.05 | 52.35 | |
| Qwen2.5-VL + BA-SFT + VanillaRLVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.05 | 51.54 | |
| Mini-o3Visual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.05 | 46.68 | |
| Qwen2.5-VL + BA-SFTVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.05 | 46.6 | |
| DeepEyesVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.05 | 45.57 | |
| ChainOfFocusVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.05 | 45.25 | |
| Mini-o3Visual Bandwidth Constraint (B)=256, Tool-calling=True2026.05 | 36.75 |