Tianhua Xia

CodeQuant: Unified Clustering and Quantization for Enhanced Outlier Smoothing in Low-Precision Mixture-of-Experts

STAR: Speculative Decoding with Searchable Drafting and Target-Aware Refinement for Multimodal Generation

FoSAM: Focus-Oriented Adaptive Token Sampling for Efficient Segment Anything in Augmented Reality

DREAM: Drafting with Refined Target Features and Entropy-Adaptive Cross-Attention Fusion for Multimodal Speculative Decoding