List your company

Hallucination Benchmark on POPE-R

88.6AccuracyV-STAR
81.8483.59585.3587.105Apr 11, 2026

Evaluation results

MethodLinks
V-STARData Size=40k2026.0488.6
Vision-R1Data Size=210k2026.0488
ThinkLite-VLData Size=11k2026.0486.9
VL-RethinkerData Size=39k2026.0485.5
VL-CogitoData Size=80k2026.0485
R1-OnevisionData Size=155k2026.0484.6
OpenVLThinkerData Size=59.2k2026.0482.4
Qwen2.5VL2026.0482.1