English
Related papers

Related papers: Toward Inference-optimal Mixture-of-Expert Large L…

200 papers

Large language model (LLM) scaling laws are empirical formulas that estimate changes in model quality as a result of increasing parameter count and training data. However, these formulas, including the popular Deepmind Chinchilla scaling…

Machine Learning · Computer Science 2025-04-15 Nikhil Sardana , Jacob Portes , Sasha Doubov , Jonathan Frankle

We develop a theory of generalization and scaling for Mixture-of-Experts (MoE) Transformers that cleanly separates \emph{active} per-input capacity from routing combinatorics. By conditioning on fixed routing patterns and union-bounding…

Machine Learning · Computer Science 2026-04-13 Mansour Zoubeirou a Mayaki

While the scaling laws of large language models (LLMs) training have been extensively studied, optimal inference configurations of LLMs remain underexplored. We study inference scaling laws (aka test-time scaling laws) and compute-optimal…

Artificial Intelligence · Computer Science 2025-03-04 Yangzhen Wu , Zhiqing Sun , Shanda Li , Sean Welleck , Yiming Yang

We present MoE-MLA-RoPE, a novel architecture combination that combines Mixture of Experts (MoE) with Multi-head Latent Attention (MLA) and Rotary Position Embeddings (RoPE) for efficient language modeling. Our approach addresses the…

Artificial Intelligence · Computer Science 2025-08-05 Sushant Mehta , Raj Dandekar , Rajat Dandekar , Sreedath Panat

Mixture-of-Experts (MoE) has emerged as a prominent architecture for scaling model size while maintaining computational efficiency. In MoE, each token in the input sequence activates a different subset of experts determined by a routing…

Computation and Language · Computer Science 2024-11-05 Chufan Shi , Cheng Yang , Xinyu Zhu , Jiahao Wang , Taiqiang Wu , Siheng Li , Deng Cai , Yujiu Yang , Yu Meng

In the era of Large Language Models (LLMs), Mixture-of-Experts (MoE) architectures offer a promising approach to managing computational costs while scaling up model parameters. Conventional MoE-based LLMs typically employ static Top-K…

Computation and Language · Computer Science 2024-10-16 Tongtian Yue , Longteng Guo , Jie Cheng , Xuange Gao , Jing Liu

The advent of Large Language Models (LLMs) has ushered in a new era of artificial intelligence, with the potential to transform various sectors through automation and insightful analysis. The Mixture of Experts (MoE) architecture has been…

Machine Learning · Computer Science 2024-10-22 Xurui Li , Juanjuan Yao

Mixture-of-Experts (MoE) based Large Language Models (LLMs) have achieved superior performance, yet the massive memory overhead caused by storing multiple expert networks severely hinders their practical deployment. Singular Value…

Machine Learning · Computer Science 2026-02-13 Zhendong Mi , Yixiao Chen , Pu Zhao , Xiaodong Yu , Hao Wang , Yanzhi Wang , Shaoyi Huang

Sparsely Mixture of Experts (MoE) has received great interest due to its promising scaling capability with affordable computational overhead. MoE converts dense layers into sparse experts, and utilizes a gated routing network to make…

Computation and Language · Computer Science 2022-07-20 Yuan Xie , Shaohan Huang , Tianyu Chen , Furu Wei

Mixture of Experts (MoE) LLMs, characterized by their sparse activation patterns, offer a promising approach to scaling language models while avoiding proportionally increasing the inference cost. However, their large parameter sizes…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-04-15 Yichao Yuan , Lin Ma , Nishil Talati

With the advancement of serverless computing, running machine learning (ML) inference services over a serverless platform has been advocated, given its labor-free scalability and cost effectiveness. Mixture-of-Experts (MoE) models have been…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-01-10 Mengfan Liu , Wei Wang , Chuan Wu

Test-Time Scaling (TTS) enhances the reasoning ability of large language models (LLMs) by allocating additional computation during inference. However, existing approaches primarily rely on output-level sampling while overlooking the role of…

Artificial Intelligence · Computer Science 2025-09-29 Yixuan Han , Fan Ma , Ruijie Quan , Yi Yang

Mixture-of-Experts (MoE) enables efficient scaling of large language models by activating only a subset of experts per input token. However, deploying MoE-based models incurs significant memory overhead due to the need to retain all experts…

Machine Learning · Computer Science 2026-02-24 Geng Zhang , Yuxuan Han , Yuxuan Lou , Yiqi Zhang , Wangbo Zhao , Yang You

Artificial intelligence (AI) has achieved astonishing successes in many domains, especially with the recent breakthroughs in the development of foundational large models. These large models, leveraging their extensive training data, provide…

Machine Learning · Computer Science 2026-01-27 Siyuan Mu , Sen Lin

Despite recent advancements of large language models (LLMs), optimally predicting the model size for LLM pretraining or allocating optimal resources still remains a challenge. Several efforts have addressed the challenge by proposing…

Machine Learning · Computer Science 2026-02-11 Md Arafat Hossain , Xingfu Wu , Valerie Taylor , Ali Jannesari

The mixture of experts (MoE) model is a sparse variant of large language models (LLMs), designed to hold a better balance between intelligent capability and computational overhead. Despite its benefits, MoE is still too expensive to deploy…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-04-23 Haodong Wang , Qihua Zhou , Zicong Hong , Song Guo

Mixture of experts (MoE) architectures have become a cornerstone for scaling up and are a key component in most large language models such as GPT-OSS, DeepSeek-V3, Llama-4, and Gemini-2.5. However, systematic research on MoE remains…

Computation and Language · Computer Science 2026-02-11 Nam V. Nguyen , Thong T. Doan , Luong Tran , Van Nguyen , Quang Pham

The Mixture-of-Experts (MoE) architecture enables a significant increase in the total number of model parameters with minimal computational overhead. However, it is not clear what performance tradeoffs, if any, exist between MoEs and…

Scaling the capacity of language models has consistently proven to be a reliable approach for improving performance and unlocking new capabilities. Capacity can be primarily defined by two dimensions: the number of model parameters and the…

Machine Learning · Computer Science 2025-07-04 Samira Abnar , Harshay Shah , Dan Busbridge , Alaaeldin Mohamed Elnouby Ali , Josh Susskind , Vimal Thilak

Mixture-of-Experts (MoE) layers have emerged as an important tool in scaling up modern neural networks by decoupling total trainable parameters from activated parameters in the forward pass for each token. However, sparse MoEs add…

Machine Learning · Computer Science 2026-05-22 Tianze Jiang , Blake Bordelon , Cengiz Pehlevan , Boris Hanin