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

Multi-modal Question Answering on MMMU

82.3Accuracy

Qwen3.5-27B

23.74838.94954.1569.351May 26, 2025Jul 26, 2025Sep 25, 2025Nov 25, 2025Jan 25, 2026Mar 27, 2026May 27, 2026
Updated 6d ago

Evaluation Results

MethodLinks
2026.04
82.3
2026.04
80.6
2026.04
79
2026.04
78.7
2026.04
78.1
2026.05
69.6
2026.05
68.1
2026.05
68.1
2026.05
62.7
2026.05
55
2026.05
54.7
2026.05
54.6
2025.06
54.14
2025.06
53.54
2025.06
53.43
2026.05
53.4
2025.06
53.08
2026.05
53
2025.06
52.6
2025.06
52.41
2025.06
51.85
2025.06
51.18
2025.06
51.13
2025.06
51.11
51
2025.10
51
2025.10
51
2025.10
50.89
2025.10
50.4
50.3
2025.06
50.15
2025.06
49.75
2025.05
49.3
2025.10
48.9
2026.05
48.6
2026.05
48.6
2025.10
47.4
2025.10
47.3
2025.10
47.1
2025.06
46.63
2025.10
46.5
2025.10
46
2025.10
45.1
2025.06
42.79
2026.05
42.7
2025.10
41.6
2026.05
41.1
2025.10
40.4
2025.05
39.33
2026.05
39.3
2026.05
39.1
2026.05
39
2025.10
38.8
2025.05
38
2025.10
37.2
2025.10
36.9
2025.10
36.9
2025.05
36.2
2025.10
36.1
2025.05
35.9
2025.10
35.6
2025.06
35.13
2025.10
35
2025.06
34.51
2025.06
34.48
2025.05
34.4
2025.06
34.39
2025.06
34.29
2025.05
34.22
2026.05
33.7
2025.05
33.67
2025.06
33.51
2025.06
33.1
2026.05
32.6
2026.05
32.2
2026.05
32
2025.05
30.33
2025.05
28.7
2025.05
28.7
2026.05
27.9
2025.05
27.6
2025.05
27.4
2025.05
26