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

General VQA on MMBench en (dev)

90.3Accuracy

K2.5

82.81284.75686.788.644Jul 10, 2025Sep 6, 2025Nov 3, 2025Dec 31, 2025Feb 27, 2026Apr 26, 2026Jun 23, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
90.3
2026.06
90.2
2026.06
89.5
2026.06
89
2025.07
88.6
2026.06
87.8
2026.06
87.7
2026.06
87.1
2026.06
86.6
2026.06
86.5
2026.06
85.6
2025.07
84.4
2026.06
84
2026.06
84
2025.07
83.1