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

Action-relation hallucination evaluation on R-Bench Image

83.5Accuracy

RVE

72.309675.214878.1281.0252May 12, 2026
Updated 21d ago

Evaluation Results

MethodLinks
2026.05
83.588.13
2026.05
83.2687.76
2026.05
83.0587.83
2026.05
82.9887.76
2026.05
82.8987.76
2026.05
82.8987.53
2026.05
82.7687.38
2026.05
82.5587.2
2026.05
82.587.49
2026.05
82.4887.59
2026.05
81.8487.45
2026.05
81.7487.31
2026.05
81.6687.2
2026.05
81.486.98
2026.05
81.3887.24
2026.05
81.3686.23
2026.05
81.3586.87
2026.05
80.4786.73
2026.05
79.8186.38
2026.05
79.185.64
2026.05
78.6285.54
2026.05
78.0585.34
2026.05
77.6985.23
2026.05
74.883.22
2026.05
72.7482.26