The paper introduces two token-error supervision methods for sparse mixture-of-experts (MoE) large language models, aligning routing affinities with token-level cross-entropy loss. The first method predicts an error score per expert and uses it to adjust affinities before top‑K selection, while the second directly aligns router affinities to the model’s objective without extra heads or inference changes. Experiments on Granite and ARC‑Challenge show accuracy gains of about 2.3–2.94 percentage points over a parameter‑matched baseline, preserving the native sparse execution budget.
By Yury Nahshan, Nati Daniel, Jacob Goldberger, Yoli Shavit
arXiv:2609.09241v1 Announce Type: cross
Abstract: Mixture-of-Experts (MoE) architectures have emerged as a powerful paradigm for scaling model capacity while preserving efficient inference in large f...
By Dohyeon Kim, Bedionita Soro, Sung Ju Hwang
The paper presents a statistical framework for Mixture-of-Experts (MoE) models, treating them as localized aggregation systems. It derives oracle risk bounds that separate approximation, expert‑learning, and router‑estimation errors for both dense and sparse routing with evolving experts. The authors also analyze how sparse Top‑K routing balances computational cost with performance, interpret gating geometrically, and explain how shared experts can capture common predictive structure while allowing routed experts to focus on local residuals.
By Siyuan He, Bokai Yang, Jie Hu, Ziwen Gao, Yuhong Yang
The paper introduces Ban&Pick, a post‑training, plug‑and‑play routing strategy for Sparse Mixture‑of‑Experts large language models. It identifies and reinforces a small group of highly influential experts while dynamically pruning redundant ones, leading to accuracy gains across math, code, and reasoning benchmarks. Experiments on DeepSeek and Qwen3 show notable performance improvements and a 1.25× inference speedup without retraining or architectural changes.
By Yuanteng Chen, Peisong Wang, Yuantian Shao, Nanxin Zeng, Chang Xu, Jian Cheng
arXiv:2608. 10605v1 Announce Type: cross Abstract: In large-scale pretraining, the algorithm, architecture, and systems decisions are conventionally made in disconnected stages.
By Soumajyoti Sarkar, Yuxin Tang, Sheng Zha
arXiv:2608. 03457v1 Announce Type: new Abstract: Diffusion language models (dLLMs) offer an alternative to autoregressive (AR) language modeling, yet the scaling behavior of Mixture-of-Experts (MoE) dLLMs remains poorly understood.
By Fengqi Zhu, Shaoxuan Xu, Jingyang Ou, Zebin You, Yipeng Xing, Huabin Liu, Xiaolu Zhang, Jun Zhou, Zhenzhong Lan, Yankai Lin, Wayne Xin Zhao, Jianguo Li, Chongxuan Li, Ji-Rong Wen
arXiv:2606. 01509v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) models scale by activating only a small subset of experts per token.
By Heng Zhao, Zilei Shao, Guy Van den Broeck, Zhe Zeng
arXiv:2606. 17952v1 Announce Type: cross Abstract: Sparse Mixture-of-Experts (MoE) architectures enable scaling LLM parameters under a fixed inference budget by activating only a small subset of experts via top-$k$ routing.
By Miko{\l}aj Zasada, {\L}ukasz Struski, Jacek Tabor, Marcin Kurdziel
arXiv:2607. 12696v1 Announce Type: cross Abstract: Sparse Mixture-of-Experts (MoE) models have become an important approach for scaling Large Language Models (LLMs), but their inference efficiency depends strongly on expert activation patterns.
By Jincheng Xie, Runheng Liu, Heyan Huang, Yawen Ling, Hanbin Dai, Yu Zheng, Wen Hu
arXiv:2602. 06154v2 Announce Type: replace Abstract: Mixture-of-Experts (MoE) models scale large language models efficiently by sparsely activating experts, but once an expert is selected, it is executed fully.
By Nurbek Tastan, Stefanos Laskaridis, Karthik Nandakumar, Samuel Horvath
arXiv:2608. 10392v1 Announce Type: new Abstract: Mixture-of-experts (MoE) models have recently moved beyond routing a fixed number of complete experts.
By Gongli Zhang, Zhulin Liu, C. L. Philip Chen
In large-scale pretraining, the algorithm, architecture, and systems decisions are conventionally made in disconnected stages. A scaling law stage selects an architecture and training recipe, optimizing loss under compute constraints, and a separate systems stage then optimizes the implementation for hardware efficiency.