List your company

Hallucination Detection on VQA-Med Open-Ended 2019 (test)

78.14AUCVASE + V-Loop
50.746457.858264.9772.0818Jan 26, 2026

Evaluation results

MethodLinks
VASE + V-LoopBackbone=InternVL3-8B [20]2026.0178.1461.57
RadFlag + V-LoopBackbone=InternVL3-8B [20]2026.0176.1458.23
VASE + V-LoopBackbone=Lingshu-7B [40]2026.0176.0168.6
SE + V-LoopBackbone=InternVL3-8B [20]2026.0175.658.99
VASE + V-LoopBackbone=MedGemma-4B-it [34]2026.0173.5154.67
RadFlag + V-LoopBackbone=Lingshu-7B [40]2026.0173.4266.43
SE + V-LoopBackbone=Lingshu-7B [40]2026.0172.5166.33
VASEBackbone=InternVL3-8B [20]2026.0172.3554.67
RadFlag + V-LoopBackbone=MedGemma-4B-it [34]2026.0171.1153.89
VASEBackbone=MedGemma-4B-it [34]2026.0170.9251.46
SE + V-LoopBackbone=MedGemma-4B-it [34]2026.0169.0251.93
VASEBackbone=Lingshu-7B [40]2026.0168.0261.23
RadFlagBackbone=InternVL3-8B [20]2026.0167.6953.58
SEBackbone=InternVL3-8B [20]2026.0166.1852.19
RadFlagBackbone=MedGemma-4B-it [34]2026.0164.9136.42
SEBackbone=MedGemma-4B-it [34]2026.0161.8847.02
RadFlagBackbone=Lingshu-7B [40]2026.0155.2440.67
SEBackbone=Lingshu-7B [40]2026.0151.839.39