Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Object Hallucination Evaluation on GQA Popular split

86.07Accuracy

MESA

71.364475.18227982.8178Apr 9, 2026Apr 22, 2026May 6, 2026May 19, 2026Jun 2, 2026Jun 15, 2026Jun 29, 2026
Updated 26d ago

Evaluation Results

MethodLinks
2026.04
86.0784.89
2026.04
85.4381.35
2026.04
85.3184.15
2026.06
84.9784.95
2026.06
84.3384.82
2026.04
83.6781.31
2026.06
83.4384.39
2026.04
82.9383.91
2026.04
82.8683.7
2026.06
82.882.72
2026.06
82.6781.89
2026.04
81.8382.23
2026.04
81.7382.39
2026.06
81.282.05
2026.04
81.1682
2026.04
80.877.89
2026.06
80.7780.15
2026.04
80.7382.24
2026.04
78.5678.79
2026.04
77.6380.13
2026.06
77.5779.4
2026.06
75.9379.62
2026.06
74.878.41
2026.04
74.7666.66
2026.04
74.7166.54
2026.06
74.3778.97
2026.06
73.7777.34
2026.04
73.377.93
2026.06
72.7377.6
2026.06
72.2775.97
2026.06
71.9376.88