Xinyin Ma

10

Papers

387

Total Citations

Papers (10)

DeepCache: Accelerating Diffusion Models for Free

dKV-Cache: The Cache for Diffusion Language Models

Thinkless: LLM Learns When to Think

Auto-Controlled Image Perception in MLLMs via Visual Perception Tokens

DepGraph: Towards Any Structural Pruning

TinyFusion: Diffusion Transformers Learned Shallow

Collaborative Decoding Makes Visual Auto-Regressive Modeling Efficient

Diffusion Model is Effectively Its Own Teacher

Structural Pruning for Diffusion Models

LLM-Pruner: On the Structural Pruning of Large Language Models