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

Complex Scene Reasoning on SEEDBench

77.03Accuracy (All)

SceneAlign

56.80262.053567.30572.5565Jan 9, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.01
77.0379.89
2026.01
76.5479.12
2026.01
75.1778.46
2026.01
74.876.93
2026.01
74.5577.12
2026.01
74.0777.43
2026.01
73.4576.22
2026.01
72.175.05
2026.01
72.0975.6
2026.01
70.9474.88
2026.01
70.7273.67
2026.01
70.374.83
2026.01
59.874.6
2026.01
58.1273.52
2026.01
57.7172.8
2026.01
57.5872.73
2026.01
57.5872.74