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

Function Selection on TaskBench Multimedia

82.3Function Selection Acc

Raw Demonstrations

2.2223.0143.864.59Dec 18, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.12
82.3
2025.12
82.3
2025.12
81.7
2025.12
81.6
2025.12
80.4
2025.12
78.5
2025.12
74.8
2025.12
74
2025.12
73.8
2025.12
73.7
2025.12
73.4
2025.12
72.8
2025.12
69.1
2025.12
69.1
2025.12
67.6
2025.12
64.1
2025.12
58.2
2025.12
57.3
2025.12
54.3
2025.12
52.4
2025.12
51.8
2025.12
51.1
2025.12
50.8
2025.12
49.6
2025.12
47.8
2025.12
46.9
2025.12
45.5
2025.12
45.1
2025.12
42.1
2025.12
38.1
2025.12
20.2
2025.12
17.6
2025.12
12.1
2025.12
9.5
2025.12
7.3
2025.12
5.3