Xingrun Xing

DenseViG: Decoupled Energy-guided Graph Structure Refinement for Vision GNNs

PDTrim: Targeted Pruning for Prefill-Decode Disaggregation in Inference

Enhancing Generalization via Sharpness-Aware Trajectory Matching for Dataset Condensation

Spike-driven Large Language Model

MedREK: Retrieval-Based Editing for Medical LLMs with Key-Aware Prompts

EfficientLLM: Unified Pruning-Aware Pretraining for Auto-Designed Edge Language Models

SpikeLLM: Scaling up Spiking Neural Network to Large Language Models via Saliency-based Spiking

From Pixels to Tokens: Byte-Pair Encoding on Quantized Visual Modalities