中文
相关论文

相关论文: Three Phases of Expert Routing: How Load Balance E…

200 篇论文

Sparsely-activated Mixture-of-experts (MoE) models allow the number of parameters to greatly increase while keeping the amount of computation for a given token or a given sample unchanged. However, a poor expert routing strategy (e.g. one…

机器学习 · 计算机科学 2022-10-17 Yanqi Zhou , Tao Lei , Hanxiao Liu , Nan Du , Yanping Huang , Vincent Zhao , Andrew Dai , Zhifeng Chen , Quoc Le , James Laudon

Mixture-of-Experts (MoE) language models route each token to a small subset of experts, but whether the routes selected by a trained top-$k$ router are good ones is rarely evaluated directly. Holding the model fixed, we compare each…

机器学习 · 计算机科学 2026-05-11 Youngsik Yoon , Siwei Wang , Wei Chen , Jungseul Ok

The Mixture-of-Experts (MoE) technique can scale up the model size of Transformers with an affordable computational overhead. We point out that existing learning-to-route MoE methods suffer from the routing fluctuation issue, i.e., the…

机器学习 · 计算机科学 2022-04-19 Damai Dai , Li Dong , Shuming Ma , Bo Zheng , Zhifang Sui , Baobao Chang , Furu Wei

Mixture-of-Experts (MoE) models can scale parameter capacity by routing each token to a subset of experts through a learned gate function. While conditional routing reduces training costs, it shifts the burden on inference memory: expert…

机器学习 · 计算机科学 2025-10-07 Rana Shahout , Colin Cai , Yilun Du , Minlan Yu , Michael Mitzenmacher

Sparse Mixture of Experts (MoE) models offer a scalable and efficient architecture for training large neural networks by activating only a subset of parameters ("experts") for each input. A learned router computes a distribution over these…

机器学习 · 计算机科学 2025-10-14 Nabil Omi , Siddhartha Sen , Ali Farhadi

This paper systematically diagnoses the training failure modes of Token-Choice sparse Mixture-of-Experts (MoE) on video Diffusion Transformers. Starting from a pretrained dense model of about 5 billion parameters, we convert it into an MoE…

计算机视觉与模式识别 · 计算机科学 2026-05-20 Haiying Sha

Mixture-of-Experts (MoE) models improve transformer efficiency but lack a unified theoretical explanation, especially when both feed-forward and attention layers are allowed to specialize. To this end, we study the Mixture-of-Transformers…

机器学习 · 计算机科学 2025-11-03 Hongbo Li , Qinhang Wu , Sen Lin , Yingbin Liang , Ness B. Shroff

Scaling the size of a model enhances its capabilities but significantly increases computation complexity. Mixture-of-Experts models (MoE) address the issue by allowing model size to scale up without substantially increasing training or…

计算与语言 · 计算机科学 2024-08-30 Zhenpeng Su , Zijia Lin , Xue Bai , Xing Wu , Yizhe Xiong , Haoran Lian , Guangyuan Ma , Hui Chen , Guiguang Ding , Wei Zhou , Songlin Hu

Sparse Mixture-of-Experts (MoE) models scale capacity by routing each token to a small subset of experts. However, their routers exhibit a fundamental trade-off: strong load balancing can suppress expert specialization, while aggressive…

机器学习 · 计算机科学 2026-05-12 Gleb Molodtsov , Alexander Miasnikov , Aleksandr Beznosikov

An increasing number of LLMs employ Mixture-of-Experts (MoE) architectures where the feed-forward layer is replaced by a pool of experts and each token only activates a small subset of them. During autoregressive generation, these models…

MoE facilitates the development of large models by making the computational complexity of the model no longer scale linearly with increasing parameters. The learning sparse gating network selects a set of experts for each token to be…

机器学习 · 计算机科学 2024-04-29 Peizhuang Cong , Aomufei Yuan , Shimao Chen , Yuxuan Tian , Bowen Ye , Tong Yang

The Mixture of Experts (MoE) selects a few feed-forward networks (FFNs) per token, achieving an effective trade-off between computational cost and performance. In conventional MoE, each expert is treated as entirely independent, and experts…

机器学习 · 计算机科学 2026-01-27 Shota Takashiro , Takeshi Kojima , Shohei Taniguchi , Yusuke Iwasawa , Yutaka Matsuo

The Mixtures-of-Experts (MoE) model is a widespread distributed and integrated learning method for large language models (LLM), which is favored due to its ability to sparsify and expand models efficiently. However, the performance of MoE…

机器学习 · 计算机科学 2024-05-24 Jing Li , Zhijie Sun , Xuan He , Li Zeng , Yi Lin , Entong Li , Binfan Zheng , Rongqian Zhao , Xin Chen

Mixture-of-Experts (MoE) models are a promising way to scale up model capacity without significantly increasing computational cost. A key component of MoEs is the router, which decides which subset of parameters (experts) process which…

计算机视觉与模式识别 · 计算机科学 2024-04-22 Tianlin Liu , Mathieu Blondel , Carlos Riquelme , Joan Puigcerver

Mixture-of-Experts (MoE) models mostly use a router to assign tokens to specific expert modules, activating only partial parameters and often outperforming dense models. We argue that the separation between the router's decision-making and…

计算与语言 · 计算机科学 2025-06-02 Ang Lv , Ruobing Xie , Yining Qian , Songhao Wu , Xingwu Sun , Zhanhui Kang , Di Wang , Rui Yan

Mixture-of-Experts (MoE) is an emerging technique for scaling large models with sparse activation. MoE models are typically trained in a distributed manner with an expert parallelism scheme, where experts in each MoE layer are distributed…

分布式、并行与集群计算 · 计算机科学 2024-11-26 Fahao Chen , Peng Li , Zicong Hong , Zhou Su , Song Guo

The Mixture-of-Experts (MoE) approach has demonstrated outstanding scalability in multi-task learning including low-level upstream tasks such as concurrent removal of multiple adverse weather effects. However, the conventional MoE…

计算机视觉与模式识别 · 计算机科学 2023-12-29 Rongyu Zhang , Yulin Luo , Jiaming Liu , Huanrui Yang , Zhen Dong , Denis Gudovskiy , Tomoyuki Okuno , Yohei Nakata , Kurt Keutzer , Yuan Du , Shanghang Zhang

Token-choice Mixture-of-Experts (TC-MoE) routes each token to a fixed number of experts, limiting dynamic computation allocation and requiring auxiliary losses to maintain load balance. We propose Expert Threshold (ET) routing, where each…

人工智能 · 计算机科学 2026-03-13 Hanchi Sun , Yixin Liu , Yonghui Wu , Lichao Sun

Mixture-of-Experts (MoE) offers flexible graph reasoning by combining multiple views of a graph through a learned router. We investigate routing-aware explanations for MoE graph models in malware detection using control flow graphs (CFGs).…

密码学与安全 · 计算机科学 2026-02-24 Hossein Shokouhinejad , Roozbeh Razavi-Far , Griffin Higgins , Ali. A Ghorbani

Mixture of Experts (MoE) are rising in popularity as a means to train extremely large-scale models, yet allowing for a reasonable computational cost at inference time. Recent state-of-the-art approaches usually assume a large number of…

计算机视觉与模式识别 · 计算机科学 2023-04-13 Amelie Royer , Ilia Karmanov , Andrii Skliar , Babak Ehteshami Bejnordi , Tijmen Blankevoort
‹ 上一页 1 2 3 10 下一页 ›