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

Future Video Reasoning on FutureBench 2-Hop

86.5Accuracy

FUTURE-L1

23.99640.22356.4572.677Jun 4, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
86.5
2026.06
83.9
2026.06
73.6
2026.06
73.1
2026.06
70.5
2026.06
69.9
2026.06
69.4
2026.06
68.4
2026.06
66.8
2026.06
66.8
2026.06
65.8
2026.06
65.3
2026.06
61.7
2026.06
59.6
2026.06
58
2026.06
57.3
2026.06
57
2026.06
53.9
2026.06
49.3
2026.06
47.2
2026.06
41.9
2026.06
26.4