English
Related papers

Related papers: Llama 3 Meets MoE: Efficient Upcycling

200 papers

Conventional large language models (LLMs) are equipped with dozens of GB to TB of model parameters, making inference highly energy-intensive and costly as all the weights need to be loaded to onboard processing elements during computation.…

Hardware Architecture · Computer Science 2025-07-28 Wei-Hsing Huang , Janak Sharda , Cheng-Jhih Shih , Yuyao Kong , Faaiq Waqar , Pin-Jun Chen , Yingyan , Lin , Shimeng Yu

Mixture-of-Experts (MoE) language models dramatically expand model capacity and achieve remarkable performance without increasing per-token compute. However, can MoEs surpass dense architectures under strictly equal resource constraints --…

Computation and Language · Computer Science 2026-05-19 Houyi Li , Ka Man Lo , Shijie Xuyang , Ziqi Wang , Wenzhen Zheng , Haocheng Zhang , Zhao Li , Shuigeng Zhou , Xiangyu Zhang , Daxin Jiang

As the training of giant dense models hits the boundary on the availability and capability of the hardware resources today, Mixture-of-Experts (MoE) models become one of the most promising model architectures due to their significant…

Mixture of Experts (MoE) has become a key architectural paradigm for efficiently scaling Large Language Models (LLMs) by selectively activating a subset of parameters for each input token. However, standard MoE architectures face…

Machine Learning · Computer Science 2025-05-27 Zehua Liu , Han Wu , Ruifeng She , Xiaojin Fu , Xiongwei Han , Tao Zhong , Mingxuan Yuan

Sparse Mixture-of-Experts (MoE) is a neural architecture design that can be utilized to add learnable parameters to Large Language Models (LLMs) without increasing inference cost. Instruction tuning is a technique for training LLMs to…

While large language models (LLMs) excel on generation tasks, their decoder-only architecture often limits their potential as embedding models if no further representation finetuning is applied. Does this contradict their claim of…

Computation and Language · Computer Science 2024-10-17 Ziyue Li , Tianyi Zhou

Mixture-of-Experts (MoE) models have gained popularity in achieving state-of-the-art performance in a wide range of tasks in computer vision and natural language processing. They effectively expand the model capacity while incurring a…

Distributed, Parallel, and Cluster Computing · Computer Science 2023-06-21 Haiyang Huang , Newsha Ardalani , Anna Sun , Liu Ke , Hsien-Hsin S. Lee , Anjali Sridhar , Shruti Bhosale , Carole-Jean Wu , Benjamin Lee

Recently, inspired by the concept of sparsity, Mixture-of-Experts (MoE) models have gained increasing popularity for scaling model size while keeping the number of activated parameters constant. In this study, we thoroughly investigate the…

Computation and Language · Computer Science 2024-11-26 Xiaoye Qu , Daize Dong , Xuyang Hu , Tong Zhu , Weigao Sun , Yu Cheng

Mixture-of-Experts (MoE) models enable scalable performance by activating large parameter sets sparsely, minimizing computational overhead. To mitigate the prohibitive cost of training MoEs from scratch, recent work employs upcycling,…

Machine Learning · Computer Science 2025-11-13 Qi Wang , Hanyang Peng , Yue Yu

Mixture-of-Experts (MoE) architectures have emerged as a promising approach to scale Large Language Models (LLMs). MoE boosts the efficiency by activating a subset of experts per token. Recent works show that fine-grained experts…

Computation and Language · Computer Science 2025-10-21 Zheyue Tan , Zhiyuan Li , Tao Yuan , Dong Zhou , Weilin Liu , Yueqing Zhuang , Yadong Li , Guowei Niu , Cheng Qin , Zhuyu Yao , Congyi Liu , Haiyang Xu , Boxun Li , Guohao Dai , Bo Zhao , Yu Wang

Large language models (LLMs) have demonstrated impressive capabilities in aiding developers with tasks like code comprehension, generation, and translation. Supporting multilingual programming -- i.e., coding tasks across multiple…

Programming Languages · Computer Science 2025-06-25 Yifan Zong , Yuntian Deng , Pengyu Nie

We present MegaScale-MoE, a production system tailored for the efficient training of large-scale mixture-of-experts (MoE) models. MoE emerges as a promising architecture to scale large language models (LLMs) to unprecedented sizes, thereby…

Large multimodal Mixture-of-Experts (MoEs) effectively scale the model size to boost performance while maintaining fixed active parameters. However, previous works primarily utilized full-precision experts during sparse up-cycling. Despite…

Computer Vision and Pattern Recognition · Computer Science 2026-01-08 Hongyu Wang , Jiayu Xu , Ruiping Wang , Yan Feng , Yitao Zhai , Peng Pei , Xunliang Cai , Xilin Chen

Most recent state-of-the-art (SOTA) large language models (LLMs) use Mixture-of-Experts (MoE) architectures to scale model capacity without proportional per-token compute, enabling higher-quality outputs at manageable serving costs.…

Large Language Models (LLMs) are predominantly deployed as dense transformers, where every parameter in every feed-forward block is activated for every token. While architecturally simple, this is computationally inefficient, since…

Machine Learning · Computer Science 2025-11-27 Ivan Novikov

How to reduce compute and memory requirements of neural networks (NNs) without sacrificing performance? Many recent works use sparse Mixtures of Experts (MoEs) to build resource-efficient large language models (LMs). Here we introduce…

Machine Learning · Computer Science 2023-11-22 Róbert Csordás , Kazuki Irie , Jürgen Schmidhuber

Mixture-of-Experts (MoE) models promise efficient scaling of large language models (LLMs) by activating only a small subset of experts per token, but their parallelized inference pipelines make elastic serving challenging. Existing…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-10-06 Gursimran Singh , Timothy Yu , Haley Li , Cheng Chen , Hanieh Sadri , Qintao Zhang , Yu Zhang , Ying Xiong , Yong Zhang , Zhenan Fan

Mixture-of-Experts (MoE) Large Language Models (LLMs) face a trilemma of load imbalance, parameter redundancy, and communication overhead. We introduce a unified framework based on dynamic expert clustering and structured compression to…

Computation and Language · Computer Science 2026-02-06 Peijun Zhu , Ning Yang , Baoliang Tian , Jiayu Wei , Weihao Zhang , Haijun Zhang , Pin Lv

We present Marco-MoE, a suite of fully open multilingual sparse Mixture-of-Experts (MoE) models. Marco-MoE features a highly sparse design in which only around 5\% of the total parameters are activated per input token. This extreme…

Computation and Language · Computer Science 2026-04-29 Fan Jiang , Yu Zhao , Chenyang Lyu , Tianqi Shi , Yichao Du , Feihu Jiang , Longyue Wang , Weihua Luo

Mixture of Experts (MoE) LLMs have recently gained attention for their ability to enhance performance by selectively engaging specialized subnetworks or "experts" for each input. However, deploying MoEs on memory-constrained devices remains…