paper-with-me

홈 › Papers

Mixture of Tokens: Continuous MoE through Cross-Example Aggregation

2023-10-24 · Szymon Antoniak, Michał Krutul, Maciej Pióro, Jakub Krajewski, Jan Ludziejewski, Kamil Ciebiera, Krystian Król, Tomasz Odrzygóźdź, Marek Cygan, Sebastian Jaszczur

Mixture of Experts (MoE) models based on Transformer architecture are pushing the boundaries of language and vision tasks. The allure of these models lies in their ability to substantially increase the parameter count without a corresponding increase in FLOPs. Most widely adopted MoE models are discontinuous with respect to their parameters - often referred to as sparse. At the same time, existing continuous MoE designs either lag behind their sparse counterparts or are incompatible with autoregressive decoding. Motivated by the observation that the adaptation of fully continuous methods has been an overarching trend in deep learning, we develop Mixture of Tokens (MoT), a simple, continuous architecture that is capable of scaling the number of parameters similarly to sparse MoE models. Unlike conventional methods, MoT assigns mixtures of tokens from different examples to each expert. This architecture is fully compatible with autoregressive training and generation. Our best models not only achieve a 3x increase in training speed over dense Transformer models in language pretraining but also match the performance of state-of-the-art MoE architectures. Additionally, a close connection between MoT and MoE is demonstrated through a novel technique we call transition tuning.

📄 PDF Abstract BibTeX arXiv:2310.15961

Code (1)

llm-random/llm-random 공식 구현 pytorch

Tasks

Language ModellingLarge Language ModelMixture-of-Experts

Methods 이 논문이 사용한 방법론

SPEED The monocular depth estimation (MDE) is the task of estimating depth from a single frame. This information is an essential knowledge in many computer vision tasks such as scene…
Multi-Head Attention 설명 없음
Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Absolute Position Encodings Absolute Position Encodings are a type of position embeddings for [Transformer-based models] where positional encodings are…
Adam 설명 없음

Similar Papers 제목 키워드 기반

MixAR: Mixture Autoregressive Image Generation

2025-11-15 · Jinyuan Hu, Jiayou Zhang, Shaobo Cui, Kun Zhang 외 arxiv

Autoregressive (AR) approaches, which represent images as sequences of discrete tokens from a finite codebook, have achieved remarkable success in image generation. However, the quantization process and the limited codeb…

Computational EfficiencyImage Generation

Soft Tokens, Hard Truths

2025-09-23 · Natasha Butt, Ariel Kwiatkowski, Ismail Labiad, Julia Kempe 외 arxiv

The use of continuous instead of discrete tokens during the Chain-of-Thought (CoT) phase of reasoning LLMs has garnered attention recently, based on the intuition that a continuous mixture of discrete tokens could simula…

Reinforcement Learning

SiRA: Sparse Mixture of Low Rank Adaptation

2023-11-15 · Yun Zhu, Nevan Wichers, Chu-Cheng Lin, Xinyi Wang 외

Parameter Efficient Tuning has been an prominent approach to adapt the Large Language Model to downstream tasks. Most previous works considers adding the dense trainable parameters, where all parameters are used to adapt…

Language ModelingLanguage ModellingLarge Language Model

FRAME: Learning the Adaptation Domain with a Mixture of Fractional-Fourier Experts

2026-06-30 · Tom Saliencro, Maya Lindqvist, Rohan Desai, Priya Nair 외 arxiv

Parameter-efficient fine-tuning (PEFT) reparameterizes weight updates in a fixed basis: low-rank adapters operate in the spatial domain, while a recent line of spectral methods operates in a fixed Fourier domain. We argu…

parameter-efficient fine-tuning

Upcycling Large Language Models into Mixture of Experts

2024-10-10 · Ethan He, Abhinav Khattar, Ryan Prenger, Vijay Korthikanti 외

Upcycling pre-trained dense language models into sparse mixture-of-experts (MoE) models is an efficient approach to increase the model capacity of already trained models. However, optimal techniques for upcycling at scal…

Mixture-of-ExpertsMMLU