7 papers
xPress: Parallel Refinement for Diffusion Drafters in Speculative Decoding
Zheng Wang, Davis Wertheimer, Yu Chin Fabian Lim +4
Block-diffusion drafters like dFlash generate an entire block of draft tokens in a single forward pass, drastically reducing the overhead of multiple-token drafting in speculative…
PRESTO: Prefix-Aligned Tree Drafting for Diffusion Speculative Decoding
Zheng Wang, Zhifan Ye, Qi Cheng +8
Diffusion Large Language Models (dLLMs) have emerged as a promising alternative to autoregressive (AR) LLMs, generating tokens in parallel. This makes them effective draft models f…
Slow-Fast Policy Optimization: Reposition-Before-Update for LLM Reasoning
Ziyan Wang, Zheng Wang, Xingwei Qu +5
Reinforcement learning (RL) has become central to enhancing reasoning in large language models (LLMs). Yet on-policy algorithms such as Group Relative Policy Optimization (GRPO) of…
When Linear Attention Meets Autoregressive Decoding: Towards More Effective and Efficient Linearized Large Language Models
Haoran You, Yichao Fu, Zheng Wang +2
Autoregressive Large Language Models (LLMs) have achieved impressive performance in language tasks but face two significant bottlenecks: (1) quadratic complexity in the attention m…
Model Tells You Where to Merge: Adaptive KV Cache Merging for LLMs on Long-Context Tasks
Zheng Wang, Boxiao Jin, Zhongzhi Yu +1
How to efficiently serve Large Language Models (LLMs) has become a pressing issue because of their huge computational cost in their autoregressive generation process. To mitigate c…
Unveiling and Harnessing Hidden Attention Sinks: Enhancing Large Language Models without Training through Attention Calibration
Zhongzhi Yu, Zheng Wang, Yonggan Fu +3
Attention is a fundamental component behind the remarkable achievements of large language models (LLMs). However, our current understanding of the attention mechanism, especially r…