English(EN) What We Keep of What We Forget: Replacement Operators and Routing Semantics in Training-Free MoE Expert Pruning

新研究探讨训练无关的 MoE 专家剪枝技术

作者 PulseAugur 编辑部 · [1 个来源] · 2026-11-02 05:00

一篇新研究论文探讨了在无需重新训练的情况下剪枝混合专家（MoE）模型中专家的方法。该研究调查了不同的替换算子和路由语义在移除专家时对模型性能的影响。主要发现表明，存活专家的门控尺度显著影响准确性，而一种新颖的“倾斜新手”算子在增加参数的情况下保持性能方面显示出潜力。 AI

影响引入了优化 MoE 模型的新技术，有望带来更高效、性能更佳的 AI 系统。

排序理由在 arXiv 上发表的研究论文，详细介绍了 MoE 专家剪枝的新方法。[lever_c_demoted from research: ic=1 ai=1.0]

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 1 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 在 arXiv 上发表的研究论文，详细介绍了 MoE 专家剪枝的新方法。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, model release; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: 1 days old; Coverage has settled into its steady-state source set.

完整方法见我们的编辑标准。

arXiv cs.CL TIER_1 English(EN) · Ezza Ali, Sehrish Bibi, Momina Moetesum, Faisal Shafait · 2026-11-02 05:00

What We Keep of What We Forget: Replacement Operators and Routing Semantics in Training-Free MoE Expert Pruning

arXiv:2610.41366v1 Announce Type: cross Abstract: 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. Holding the mask fixed, we compare repla…