Kun Gai

8

Papers

135

Total Citations

Papers (8)

Improving Video Generation with Human Feedback

LEARN: Knowledge Adaptation from Large Language Model to Recommendation for Practical Industrial Application

Who You Are Matters: Bridging Interests and Social Roles via LLM-Enhanced Logic Recommendation

Granularity-Adaptive Spatial Evidence Tokenization for Video Question Answering

Towards Precise Scaling Laws for Video Diffusion Transformers

Video-LaVIT: Unified Video-Language Pre-training with Decoupled Visual-Motional Tokenization

Imbalance in Balance: Online Concept Balancing in Generation Models

FullDiT: Video Generative Foundation Models with Multimodal Control via Full Attention