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

Multi-modal Reasoning on MMStar

63.78Accuracy

SINKTRACK

23.334433.834744.33554.8353Apr 11, 2026Apr 12, 2026Apr 14, 2026Apr 15, 2026Apr 17, 2026Apr 18, 2026Apr 20, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.04
63.7877.86
2026.04
61.675.81
2026.04
60.5675.09
2026.04
57.1-
2026.04
55.8471.29
2026.04
55.4-
2026.04
53.0469.11
2026.04
51.3667.52
2026.04
48.1-
2026.04
47.2461.38
2026.04
47.1-
2026.04
46.9-
2026.04
44.8-
2026.04
44.5860.89
2026.04
43.4-
2026.04
40.4-
2026.04
40.3-
2026.04
40.2-
2026.04
39.4-
2026.04
38.5-
2026.04
37.4-
2026.04
37.2954.24
2026.04
35.4752.17
2026.04
35.3-
2026.04
35.2450.68
2026.04
33.4-
2026.04
33.2-
2026.04
24.8939.54