논문해외
What We Keep of What We Forget: Replacement Operators and Routing Semantics in Training-Free MoE Expert Pruning
HF Daily Papers
Training-free expert pruning makes two decisions: which experts to remove, and what the layer computes when its router still selects a removed expert. Published methods change both at once.
원문의 저작권은 HF Daily Papers에 있습니다. 위 발췌는 원문 확인을 돕기 위한 짧은 인용입니다.