← 研踪 PaperTrailCVF · CVPR 2026:cvf:bc8d4d752a062c7e61bde9e800a5b9c3 · 记录 #1
ERMoE: Eigen-Reparameterized Mixture-of-Experts for Stable Routing and Interpretable Specialization
Cheng, AnzheDuan, ShukaiLi, ShixuanYin, ChenzhongCheng, MingxiPing, HengChattopadhyay, TamoghnaThomopoulos, Sophia I.Nazarian, ShahinThompson, PaulBogdan, Paul 来源日期（仅到年份） 2026
来源记录日期 2026
雷达观测 2026-11-05 03:59
摘要
Mixture-of-Experts (MoE) models expand capacity via sparse expert activation, but routing logits can misalign with expert structure (unstable routing, underutilization) and load imbalance can create stragglers. Auxiliary load-balancing losses reduce disparity but often weaken specialization and downstream accuracy. We propose ERMoE, a sparse MoE transformer that reparameterizes each expert in a learned orthonormal eigenbasis and routes with an Eigenbasis Score (cosine similarity between token features and an expert basis) instead of learned gating logits. By tying assignments to each expert's representation space, ERMoE stabilizes utilization, improves interpretability, and removes explicit balancing losses and their gradient interference. ERMoE reaches state-of-the-art results on ImageNet and image-text retrieval (COCO, Flickr30K) with flatter expert loads. A 3D MRI variant (ERMoE-ba) improves brain age prediction by over 7% and yields anatomically interpretable expert specializations.
论文发现出处
这些页面提供论文线索，不代表同行评审或质量认证。
观测到的版本
平台记录号代表本站观测到的元数据变化，不是作者修订号。来源日期仅到月份或年份时，按该月或该年首日存档。
记录 #1 · 2026 · 最新观测版本
ERMoE: Eigen-Reparameterized Mixture-of-Experts for Stable Routing and Interpretable Specialization
Mixture-of-Experts (MoE) models expand capacity via sparse expert activation, but routing logits can misalign with expert structure (unstable routing, underutilization) and load imbalance can create stragglers. Auxiliary load-balancing losses reduce disparity but often weaken specialization and downstream accuracy. We propose ERMoE, a sparse MoE transformer that reparameterizes each expert in a learned orthonormal eigenbasis and routes with an Eigenbasis Score (cosine similarity between token features and an expert basis) instead of learned gating logits. By tying assignments to each expert's representation space, ERMoE stabilizes utilization, improves interpretability, and removes explicit balancing losses and their gradient interference. ERMoE reaches state-of-the-art results on ImageNet and image-text retrieval (COCO, Flickr30K) with flatter expert loads. A 3D MRI variant (ERMoE-ba) improves brain age prediction by over 7% and yields anatomically interpretable expert specializations.
查看此版本解读 →查看此版本原文 ↗