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

Cooking

Benchmarks

Task NameDataset NameSOTA ResultTrend
Controllable video generationCooking 50 real-world videos (test)
PSNR16.44
6
Interleaved generationCooking-200 Text Input
T-Com4.02
5
Interleaved generationCooking-200
T-Com4.24
5
Cross-task GeneralizationCooking (test)
Similarity0.6889
4
Action AlignmentCooking 2 (test)
Midpoint Score10.6
4
Human Preference EvaluationCooking
Step Faithfulness Win Rate94
3
Showing 6 of 6 rows