Xinting Hu

SPEED: Scalable, Precise, and Efficient Concept Erasure for Diffusion Models

Easier Painting Than Thinking: Can Text-to-Image Models Set the Stage, but Not Direct the Play?

On the Generalization of SFT: A Reinforcement Learning Perspective with Reward Rectification

Memory Forcing: Spatio-Temporal Memory for Consistent Scene Generation on Minecraft

LayerCake: Token-Aware Contrastive Decoding within Large Language Model Layers

SPEED: Scalable, Precise, and Efficient Concept Erasure for Diffusion Models