Kun Gai

UniVideo: Unified Understanding, Generation, and Editing for Videos

Unified In-Context Video Editing

Mitigating Noise Shift in Denoising Generative Models with Noise Awareness Guidance

Improving Autoregressive Video Modeling with History Understanding

Denoising Neural Reranker for Recommender Systems

Group Relative Attention Guidance for Image Editing

DiffMoE: Dynamic Token Selection for Scalable Diffusion Transformers

Stabilizing Knowledge, Promoting Reasoning: Dual-Token Constraints for RLVR

ALM-MTA: Front-Door Causal Multi-Touch Attribution Method for Creator-Ecosystem Optimization

RelightMaster: Precise Video Relighting with Multi-plane Light Images

GoalRank: Group-Relative Optimization for a Large Ranking Model

Generative Auto-Bidding in Large-Scale Auctions via Diffusion Completer-Aligner

A Reason-then-Describe Instruction Interpreter for Controllable Video Generation

Boosting Latent Diffusion Models via Semantic-Disentangled VAE

Scaling Image and Video Generation via Test-Time Evolutionary Search

Bridging Modalities for Forgery Detection via Learnable Representations with Query-Guided Contrastive Learning

FoleyGenEx: Unified Video-to-Audio Generation with Multi-Modal Control, Temporal Alignment, and Semantic Precision

Interpreting Any Condition to Caption for Controllable Video Generation

OneSearch: A Preliminary Exploration of the Unified End-to-End Generative Framework for E-commerce Search

Sample by Step, Optimize by Chunk: Chunk-Level GRPO for Text-to-Image Generation

Attention as a Compass: Efficient Exploration for Process-Supervised RL in Reasoning Models

Score Augmentation for Diffusion Models

VideoCanvas: Unified Video Completion from Arbitrary Spatiotemporal Patches via In-Context Conditioning

PR-VTON: Enhancing Detail Fidelity in Virtual Try-On with Refined Positional Encoding

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

Learning Cascade Ranking as One Network

Fading to Grow: Growing Preference Ratios via Preference Fading Discrete Diffusion for Recommendation

Improving Video Generation with Human Feedback

RecFlow: An Industrial Full Flow Recommendation Dataset

Making Transformer Decoders Better Differentiable Indexers

Learning Time-Aware Causal Representation for Model Generalization in Evolving Domains

Multi-Epoch Learning with Data Augmentation for Deep Click-Through Rate Prediction

Generate explorative goals with large language model guidance