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

Compositional Reasoning on GenAI-Bench (test)

83.79Spatial Score

CDG

69.084472.902276.7280.5378Mar 11, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.03
83.7980.2483.5477.56
2026.03
83.2680.1983.4177.36
2026.03
80.6976.0678.7473.13
2026.03
79.8474.8678.2671.53
2026.03
79.6673.775.172.21
2026.03
79.5873.5475.0871.94
2026.03
78.974.1375.6370.23
2026.03
78.6674.0877.2270.74
2026.03
78.5574.0576.9870.11
2026.03
78.0973.175.8969.39
2026.03
77.5673.4776.1771.55
2026.03
77.4772.9775.3971.13
2026.03
77.4272.5874.8770.81
2026.03
77.0772.8374.571.47
2026.03
76.3471.4372.867.73
2026.03
75.6471.0272.3866.17
2026.03
72.2669.5170.4665.52
2026.03
69.6566.467.0964.8