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

Complex Reasoning on vcrbench

53.1Accuracy

VideoLatent-7B (Ours)

23.56431.23238.946.568Mar 18, 2026Apr 3, 2026Apr 19, 2026May 5, 2026May 21, 2026Jun 6, 2026Jun 22, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
53.1
2026.06
52.5
2026.06
52.5
2026.06
51.4
2026.06
50.8
2026.06
50.6
2026.06
50.6
2026.06
50.4
2026.06
49.6
2026.06
49.2
2026.06
49
2026.06
48.6
2026.06
48
2026.06
46.9
2026.06
46.7
2026.06
46.3
2026.06
45.3
2026.03
35.6
2026.03
34.2
2026.03
31.5
2026.03
30.1
2026.06
25.5
2026.06
25.1
2026.06
24.7