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

Multimodal Understanding on MMMU (Accuracy)

56.8Accuracy

GPT-4V

21.12830.38939.6548.911Jan 13, 2025Mar 23, 2025May 31, 2025Aug 9, 2025Oct 17, 2025Dec 25, 2025Mar 5, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2025.01
56.8
2025.01
50.7
2025.01
49.7
2026.03
46.7
2026.03
46.7
2026.03
46.7
2026.03
46.6
2026.03
46.2
2026.03
45.6
2026.03
44.4
2026.03
44.4
2026.03
43.3
2026.03
43.3
2026.03
43.3
2026.03
43.3
2026.03
43.3
2026.03
43.3
2026.03
43.3
2026.03
43.3
2026.03
42.6
2026.03
42.6
2026.03
42.2
2026.03
41.9
2026.03
41.9
2026.03
41.2
2026.03
40.9
2026.03
40
2026.03
38.7
2025.01
38.6
2026.03
37.8
2025.01
36.4
2025.01
35.8
2026.03
30
2026.03
29.7
2026.03
28.9
2026.03
25.6
2026.03
25.6
2026.03
22.5