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

Vision-related Reasoning and Perception on MMBench (test)

80.38Avg Score

Rotation

20.704836.197451.6967.1826Sep 26, 2023Jan 29, 2024Jun 3, 2024Oct 6, 2024Feb 9, 2025Jun 14, 2025Oct 18, 2025
Updated 15d ago

Evaluation Results

MethodLinks
2025.10
80.38-83.8980.54---65.8480.2171.5384.76
2025.10
80.08-82.2277.19---67.6582.1566.5185.39
2025.10
77.82-80.3574.51---62.8677.9267.8284.31
2024.05
76.8----------
2024.05
76.7----------
74.450.68276.179.359.281.7----
2024.05
73.2----------
2025.10
72.99-76.6241.54---61.7773.5564.3282.06
2024.05
72.3----------
2024.05
70----------
2025.10
69.27-71.8173.42---61.1265.3858.3978.5
2023.09
6643.47662.168.655.973----
2023.09
65.244.377.964.866.553.670.6----
2023.09
62.64174.355.961.658.769.2----
61.840.574.347.966.346.272.8----
2023.09
60.233.569.653.161.850.471.7----
2023.09
59.532.472.649.362.352.267.7----
2023.09
48.322.263.339.446.836.460.6----
2023.09
38.97.445.319.2453254----
2023.09
36.215.953.628.641.82040.4----
33.921.647.422.53324.441.1----
2023.09
33.511.448.827.735.817.641.5----
2023.09
2313.632.98.928.811.228.3----