8 papers
Less Experts, Faster Decoding: Cost-Aware Speculative Decoding for Mixture-of-Experts
Jincheng Xie, Runheng Liu, Heyan Huang +4
Sparse Mixture-of-Experts (MoE) models have become an important approach for scaling Large Language Models (LLMs), but their inference efficiency depends strongly on expert activat…
AdaPLD: Adaptive Retrieval and Reuse for Efficient Model-Free Speculative Decoding
Runheng Liu, Jincheng Xie, Wen Hu +2
Speculative decoding accelerates generation by verifying multiple drafted tokens in a single target-model forward pass, reducing sequential decoding iterations. Model-free variants…
SPECTRE: Hybrid Ordinary-Parallel Speculative Serving for Resource-Efficient LLM Inference
Jincheng Xie, Yawen Ling, Qi Xiao +4
LLM serving platforms are increasingly deployed as multi-model cloud systems, where user demand is often long-tailed: a few popular large models receive most requests, while many s…
SAT-BO: Verification Rule Learning and Optimization for FraudTransaction Detection
Mao Luo, Zhi Wang, Yiwen Huang +5
Electronic payment platforms are estimated to process billions oftransactions daily, with the cumulative value of these transactionspotentially reaching into the trillions. Even a…
CAKE: Cascading and Adaptive KV Cache Eviction with Layer Preferences
Ziran Qin, Yuchen Cao, Mingbao Lin +5
Large language models (LLMs) excel at processing long sequences, boosting demand for key-value (KV) caching. While recent efforts to evict KV cache have alleviated the inference bu…
SCOOT: SLO-Oriented Performance Tuning for LLM Inference Engines
Ke Cheng, Zhi Wang, Wen Hu +3
As large language models (LLMs) are gaining increasing popularity across a wide range of web applications, it is of great importance to optimize service-level objectives (SLOs) for…