Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Hallucination Evaluation on POPE Popular v1.0 (test)

90.34Accuracy

ResDec

75.5279.367583.21587.0625Feb 1, 2026Feb 2, 2026Feb 4, 2026Feb 6, 2026Feb 8, 2026Feb 10, 2026Feb 12, 2026
Updated 3d ago

Evaluation Results

MethodLinks
2026.02
90.3489.83
2026.02
89.2388.21
2026.02
88.8987.05
2026.02
87.9288.95
2026.02
87.8485.78
2026.02
87.8485.78
2026.02
87.0185.28
2026.02
86.9787.22
2026.02
86.3586.14
2026.02
86.0384.56
2026.02
85.9785.87
2026.02
85.785.1
2026.02
85.1484.68
2026.02
85.184.6
2026.02
84.2183.5
2026.02
8382.9
2026.02
82.6583.37
2026.02
82.0183.42
2026.02
81.8682.58
2026.02
81.3480.94
2026.02
81.0880.74
2026.02
80.879.9
2026.02
80.4382.07
2026.02
79.9879.34
2026.02
79.9480.8
2026.02
79.7580.61
2026.02
79.6882.27
2026.02
78.9979.85
2026.02
78.3380.2
2026.02
78.1379.12
2026.02
76.0977.65