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

Visual Reasoning on GQA (Accuracy)

73.01Accuracy

ActiveScope

-2.920416.792336.50556.2177May 15, 2025Jul 21, 2025Sep 26, 2025Dec 3, 2025Feb 8, 2026Apr 16, 2026Jun 23, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
73.01-
2026.06
72.82-
2026.06
72.34-
2026.04
64.54-
2025.05
64.2-
2026.04
64.14-
2026.04
63.97-
2026.03
63.25-
2025.05
62.6-
2025.11
62.6-
2026.03
62.5100
2026.03
62.43-
2026.03
6299.64
2026.04
61.98-
2026.03
61.998.9
2025.05
61.9-
2025.11
61.9-
2025.05
61.7-
2025.11
61.7-
2025.11
61.5-
2025.05
61.4-
2025.11
61.3-
2026.03
61.298.8
2026.03
61.299.8
2025.05
61.2-
2026.04
61.15-
2026.03
61.197
2025.11
61.1-
2026.04
61.01-
2026.04
60.97-
2025.05
60.9-
2026.03
60.898.3
2025.11
60.8-
2025.05
60.7-
2025.11
60.6-
2025.11
60.5-
2026.03
60.197.94
2026.03
6097
2025.05
60-
2026.03
59.72-
2025.11
59.5-
2026.04
59.49-
2026.03
59.12-
2026.04
59.01-
2025.05
59-
2025.05
59-
2026.03
58.995.2
2026.03
58.996.4
2026.03
58.82-
2026.03
58.895.7
2026.03
58.76-
2025.05
58.7-
2026.03
58.67-
2026.02
58.5-
2026.02
58.5-
2026.02
58.3-
2026.02
58.3-
2026.02
58.2-
2025.05
58.2-
2026.04
58.04-
2026.02
57.9-
2026.03
57.994.6
2026.03
57.85-
2026.03
57.8-
2025.11
57.7-
2025.05
57.6-
2026.03
57.32-
2026.02
57.3-
2025.05
57.3-
2026.03
57.19-
2025.11
57-
2025.05
56.9-
2025.05
56.9-
2025.05
56.1-
2025.05
56-
2025.05
55.9-
2025.05
55.9-
2026.02
55.7-
2026.02
55.1-
2025.05
53.1-
2026.02
52.7-
2025.05
52.7-
2025.05
52.7-
2026.03
50.482.5
2026.04
50.18-
2026.03
49.879.9
2025.05
49.6-
2025.05
46.1-
2025.11
21.8-
2025.11
21.6-
2025.11
19.8-
2025.11
16.4-
2025.11
8.2-
2025.11
6.8-
2025.11
0-
2025.11
0-