List your company
ResearchFormerly WizWand

Hallucination Evaluation on POPE Overall

91.1AccuracyInternVL3-8B
82.67684.86387.0589.237Sep 26, 2025Nov 3, 2025Dec 11, 2025Jan 18, 2026Feb 25, 2026Apr 4, 2026May 12, 2026
Updated 5mo ago

Evaluation results

MethodLinks
InternVL3-8BCategory=SFT & Close-source Models, Parameter Size=8B2025.0991.1—
AdaVaR-7BCategory=Reasoning Models, Base Model=Qwen2.5-VL-7B, Parameter Size=7B, Reasoning Mode=Adaptive2025.098955.82
LLaVA-OV-7BCategory=SFT & Close-source Models, Parameter Size=7B2025.0988.4—
Chain-of-FocusCategory=Reasoning Models, Base Model=Qwen2.5-VL-7B, Parameter Size=7B2025.0988.450.94
ViGoRL-7BCategory=Reasoning Models, Base Model=Qwen2.5-VL-7B, Parameter Size=7B2025.0988.350.53
AdaVaR-3BCategory=Reasoning Models, Base Model=Qwen2.5-VL-3B, Parameter Size=3B, Reasoning Mode=Adaptive2025.0988.250.84
ViGoRL-3BCategory=Reasoning Models, Base Model=Qwen2.5-VL-3B, Parameter Size=3B2025.098845.7
DeepEyesCategory=Reasoning Models, Base Model=Qwen2.5-VL-7B, Parameter Size=7B, Reasoning Mode=Visually-grounded2025.0987.953.72
VLAA-Thinker-3BCategory=Reasoning Models, Base Model=Qwen2.5-VL-3B, Parameter Size=3B, Reasoning Mode=Text-based2025.0987.846.7
Qwen2.5-VL-7BCategory=Reasoning Models, Base Model=Qwen2.5-VL-7B, Parameter Size=7B2025.0987.850.9
Qwen2.5-VL-3BCategory=Reasoning Models, Base Model=Qwen2.5-VL-3B, Parameter Size=3B2025.0987.245.74
GPT-4oCategory=SFT & Close-source Models2025.0986.953.2
Orsta-7BCategory=Reasoning Models, Base Model=Qwen2.5-VL-7B, Parameter Size=7B2025.0986.952.04
LMM-R1Category=Reasoning Models, Base Model=Qwen2.5-VL-3B, Parameter Size=3B2025.0986.748.8
GAP (LH+PCA+DA)Latent Head=true, PCA=true, Data Augmentation=true2026.0586.3—
MM-EurekaCategory=Reasoning Models, Base Model=Qwen2.5-VL-7B, Parameter Size=7B2025.0986.352.52
Qwen2.5-VL-7BModel size=7B2026.0586.25—
VLAA-Thinker-7BCategory=Reasoning Models, Base Model=Qwen2.5-VL-7B, Parameter Size=7B, Reasoning Mode=Text-based2025.0986.254.2
LH+DA (no PCA)Latent Head=true, Data Augmentation=true, PCA=false2026.0586.02—
OVR-7BCategory=Reasoning Models, Base Model=Qwen2.5-VL-7B, Parameter Size=7B, Reasoning Mode=Text-based2025.0983.255.76
GRIT-3BCategory=Reasoning Models, Base Model=Qwen2.5-VL-3B, Parameter Size=3B2025.098345.8