List your company

Hallucination Detection on VQA-Med 2019 (All)

78.1AUCVIHD
36.489647.292358.09568.8977Jan 26, 2026Feb 14, 2026Mar 5, 2026Mar 24, 2026Apr 12, 2026May 1, 2026May 20, 2026

Evaluation results

MethodLinks
VIHDBackbone=Hulu-4B2026.0578.160.39
VASE + V-LoopBackbone=InternVL3-8B [20]2026.0176.161.76
RadFlag + V-LoopBackbone=InternVL3-8B [20]2026.0175.9260.79
VIHDBackbone=Lingshu-7B2026.0574.1864.54
SE + V-LoopBackbone=InternVL3-8B [20]2026.0174.1457.67
VASEBackbone=Hulu-4B2026.0572.9957.41
VASEBackbone=InternVL3-8B [20]2026.0172.0754.18
V-LoopBackbone=InternVL3-8B2026.0170.5555.82
RadFlagBackbone=InternVL3-8B, Uncertainty-based=true2026.0170.2555.28
RadFlagBackbone=InternVL3-8B [20]2026.0170.2555.28
MaxEntBackbone=Hulu-4B2026.0569.6455.45
VASE + V-LoopBackbone=MedGemma-4B-it [34]2026.0169.5654.9
MaxProbBackbone=Hulu-4B2026.0569.4555.39
VASEBackbone=Lingshu-7B2026.0568.8560.8
VASE + V-LoopBackbone=Lingshu-7B [40]2026.0168.868.43
VASEBackbone=MedGemma-4B-it [34]2026.0168.1944.71
AvgProbBackbone=Hulu-4B2026.0567.955
AvgEntBackbone=Hulu-4B2026.0567.7255
RadFlag + V-LoopBackbone=Lingshu-7B [40]2026.0167.2365.22
VASEBackbone=Lingshu-7B [40]2026.0167.0957.33
RadFlag + V-LoopBackbone=MedGemma-4B-it [34]2026.0167.0253.91
SEBackbone=InternVL3-8B, Uncertainty-based=true2026.0166.3546.71
SEBackbone=InternVL3-8B [20]2026.0166.3546.71
SE + V-LoopBackbone=MedGemma-4B-it [34]2026.0166.0753.77
SE + V-LoopBackbone=Lingshu-7B [40]2026.0166.0165.62
V-LoopBackbone=Lingshu-7B2026.0165.6665.03
V-LoopBackbone=MedGemma-4B-it2026.0163.2852.15
RadFlagBackbone=MedGemma-4B-it, Uncertainty-based=true2026.0162.9550.14
RadFlagBackbone=MedGemma-4B-it [34]2026.0162.9550.14
MaxEntBackbone=InternVL3-8B2026.0162.5854.87
SEBackbone=MedGemma-4B-it, Uncertainty-based=true2026.0160.9148.78
SEBackbone=MedGemma-4B-it [34]2026.0160.9148.78
AvgEntBackbone=InternVL3-8B2026.0160.1747.9
MaxEntBackbone=MedGemma-4B-it2026.0156.4234.84
RadFlagBackbone=Lingshu-7B2026.0556.0437.85
RadFlagBackbone=Lingshu-7B, Uncertainty-based=true2026.0155.9142.77
RadFlagBackbone=Lingshu-7B [40]2026.0155.9142.77
AvgEntBackbone=MedGemma-4B-it2026.0155.4235.4
MaxEntBackbone=Lingshu-7B2026.0555.1536.82
MaxProbBackbone=Lingshu-7B2026.0555.1236.78
AvgProbBackbone=Lingshu-7B2026.0555.1136.81
AvgEntBackbone=Lingshu-7B2026.0555.1136.9
MaxEntBackbone=Lingshu-7B2026.0154.441.3
AvgEntBackbone=Lingshu-7B2026.0154.3641.31
SEBackbone=Lingshu-7B2026.0553.9635.96
SEBackbone=Lingshu-7B, Uncertainty-based=true2026.0151.938.72
SEBackbone=Lingshu-7B [40]2026.0151.938.72
RadFlagBackbone=Hulu-4B2026.055022.84
SEBackbone=Hulu-4B2026.0547.4221.43
AvgProbBackbone=Lingshu-7B2026.0145.6448.61
MaxProbBackbone=Lingshu-7B2026.0145.647.82
AvgProbBackbone=MedGemma-4B-it2026.0144.4138.33
MaxProbBackbone=MedGemma-4B-it2026.0143.6435.69
AvgProbBackbone=InternVL3-8B2026.0139.7437.45
MaxProbBackbone=InternVL3-8B2026.0138.0932.09