KlingTeam/LongBanana
Viewer • Updated • 6 • 116
The Kling Team is building next-generation multimodal world models across video, audio, text, 3D, and beyond. We are continuously looking for exceptional talent to join us. Feel free to reach out!
Behind Video Generation: Observations and Insights on Multimodal Understanding from Kling
Sparsity Should Be ALLOTted by Output: Training-Free Allocation of Block-Sparse Attention for Video Diffusion Transformers