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

Spatial-relation hallucination detection on R-Bench Image

81.13Accuracy

RVE

77.479678.427379.37580.3227May 12, 2026
Updated 21d ago

Evaluation Results

MethodLinks
2026.05
81.13
2026.05
80.31
2026.05
79.67
2026.05
79.35
2026.05
79.21
2026.05
79.08
2026.05
78.21
2026.05
77.62