SPAR: Step-wise Path Dispatching and Asymmetric Re-routing for Efficient MoE Inference

Lecture notes in computer science 2026

View at publisher

Authors: Q A Zhang, X F Li, Jinzhu Kong, X Y Liu, Bin Ji, Shasha Li, Ma Jw, Jie Yu

DOI: 10.1007/978-981-92-3450-9_32

Cited by: 1

References: 7

Publication details

References 7