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

FutureBench

Benchmarks

Task NameDataset NameSOTA ResultTrend
Future Video ReasoningFutureBench Average
Accuracy85.4
22
Future Video ReasoningFutureBench Interp.
Accuracy85.1
22
Future Video ReasoningFutureBench 3-Hop
Accuracy86.6
22
Future Video ReasoningFutureBench 2-Hop
Accuracy86.5
22
Future Video ReasoningFutureBench 1-Hop
Accuracy83.2
22
Future Event PredictionFutureBench (test)
1-Hop Score80.9
20
Video Event PredictionFutureBench (test)
AVG75
3
Showing 7 of 7 rows