作者位次 #2/6ICLR 2026Poster· Deep Learning / Generative Models and Autoencoders
REAP the Experts: Why Pruning Prevails for One-Shot MoE compression
Mike Lasby · Ivan Lazarevich · Nish Sinnadurai · Sean Lie · Yani Ioannou · Vithursan Thangarasa
效率与系统OpenReview