List your company

Visual Search and Perception-intensive Reasoning on HR-Bench 8K

70ScoreStarve to Perceive
24.7636.50548.2559.995May 18, 2026

Evaluation results

MethodLinks
Starve to PerceiveVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.0570
PixelReasonerVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.0569.25
DeepEyesVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.0567.88
Qwen2.5-VL + BA-SFT + VanillaRLVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.0567.75
ChainOfFocusVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.0567.25
Qwen2.5-VL + BA-SFTVisual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.0565.88
GPT-4oVisual Bandwidth Constraint (B)=No Constrain, Reported From=Other publications2026.0563.9
Qwen2.5-VLVisual Bandwidth Constraint (B)=No Constrain, Reported From=Other publications2026.0563
Starve to PerceiveVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.0560.38
Qwen2.5-VL + BA-SFT + VanillaRLVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.0560
LLaVA-OneVisionVisual Bandwidth Constraint (B)=No Constrain, Reported From=Other publications2026.0559.8
Qwen2.5-VL + BA-SFTVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.0556.13
ChainOfFocusVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.0553.75
DeepEyesVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.0551.25
Mini-o3Visual Bandwidth Constraint (B)=No Constrain, Tool-calling=True2026.0547.75
PixelReasonerVisual Bandwidth Constraint (B)=256, Tool-calling=True2026.0543.13
Mini-o3Visual Bandwidth Constraint (B)=256, Tool-calling=True2026.0540.88
Qwen2.5-VLVisual Bandwidth Constraint (B)=2562026.0526.5