Yehui Tang

CBQ: Cross-Block Quantization for Large Language Models

Mixture of Lookup Experts

Forest-of-Thought: Scaling Test-Time Compute for Enhancing LLM Reasoning

SlimLLM: Accurate Structured Pruning for Large Language Models

Multi-Granularity Semantic Revision for Large Language Model Distillation

SAN-Diff: Structure-aware noise for super-resolution diffusion model

SpeCache: Speculative Key-Value Caching for Efficient Generation of LLMs

LLM Data Selection and Utilization via Dynamic Bi-level Optimization

Empirical Study on Enhancing Efficiency in Masked Image Modeling Pre-training

Prompt-guided Visual Perception for Efficient Training-free Video LLM

MemoryFormer : Minimize Transformer Computation by Removing Fully-Connected Layers

Star-Agents: Automatic Data Optimization with LLM Agents for Instruction Tuning

Kangaroo: Lossless Self-Speculative Decoding for Accelerating LLMs via Double Early Exiting

Connectivity-based Token Condensation for Efficient Vision Transformer