arXiv Machine Learning By Yingnan Zhao, Razvan Bunescu, Ahmed Louri, Avinash Karanth, Ke Wang

A Spatio-Temporal Expert Prefetching Framework for Efficient MoE-based LLM Inference

Read the original on arXiv Machine Learning →

arXiv:2606. 15453v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) based large language models (LLMs), such as Qwen and DeepSeek, have recently emerged as an effective approach to improving model capacity without proportionally increasing computational cost.

Summary generated by The Flow from the publisher's feed. The full article lives at arXiv Machine Learning.

Hugging Face Trending Papers
Jul 27

DraftExpert: Expansion-Aware Self-Speculative Decoding for End-Device MoE Inference

Large Mixture-of-Experts (MoE) language models are attractive for end-device deployment because only a small subset of experts is active per token, but their routed expert weights often exceed accelerator memory. We target latency-critical single-user settings where routed experts are staged on demand from CPU memory to a GPU or from Flash to a mobile NPU.