Mix-MoE: Improving Multilingual Machine Translation of Large Language Models through Mixed MoEs

Hugging Face Daily Papers Papers

Summary

Mix-MoE proposes a mixed Mixture-of-Experts framework with specialized expert groups and Fourier-transform-enhanced routing to mitigate parameter interference in multilingual machine translation, achieving significant improvements over baselines.

Large Language Models (LLMs) have shown great promise in multilingual machine translation (MT), even with limited bilingual supervision. However, fine-tuning LLMs with parallel corpora presents major challenges, namely parameter interference. To address these issues, we propose Mix-MoE, a mixed Mixture-of-Experts framework designed to train LLMs for multilingual MT. Our framework operates in two distinct stages: (1) post-pretraining with MoE on monolingual corpora, and (2) post-pretraining with MoE on parallel corpora. Crucially, we divide the MoE layers into two specialized groups: Language Model Experts (LM Experts) and Machine Translation Experts (MT Experts). LM Experts are designed to capture and retain the monolingual knowledge learned by the pre-trained LLM. MT Experts, on the other hand, are specifically trained to acquire and store bilingual translation knowledge. Furthermore, to facilitate effective interaction between these specialized experts and leverage potential underlying structural patterns in text, we introduce a routing mechanism enhanced by Fourier Transform features derived from model representations. The experimental results demonstrate that Mix-MoE excels in multilingual MT, significantly outperforming existing baselines and showing notable progress in mitigating parameter interference.
Original Article
View Cached Full Text

Cached at: 05/26/26, 02:41 AM

Paper page - Mix-MoE: Improving Multilingual Machine Translation of Large Language Models through Mixed MoEs

Source: https://huggingface.co/papers/2605.24681 Published on May 23

·

Submitted byhttps://huggingface.co/liboaccn

Bo Lion May 25

Abstract

Mix-MoE, a mixed Mixture-of-Experts framework, addresses parameter interference in multilingual machine translation by separating language modeling and translation expertise across specialized expert groups with Fourier-transform-enhanced routing.

Large Language Models (LLMs) have shown great promise inmultilingual machine translation(MT), even with limited bilingual supervision. However, fine-tuning LLMs withparallel corporapresents major challenges, namelyparameter interference. To address these issues, we propose Mix-MoE, a mixedMixture-of-Expertsframework designed to train LLMs for multilingual MT. Our framework operates in two distinct stages: (1)post-pretrainingwith MoE onmonolingual corpora, and (2)post-pretrainingwith MoE onparallel corpora. Crucially, we divide the MoE layers into two specialized groups:Language Model Experts(LM Experts) andMachine Translation Experts(MT Experts). LM Experts are designed to capture and retain the monolingual knowledge learned by the pre-trained LLM. MT Experts, on the other hand, are specifically trained to acquire and store bilingual translation knowledge. Furthermore, to facilitate effective interaction between these specialized experts and leverage potential underlying structural patterns in text, we introduce arouting mechanismenhanced byFourier Transformfeatures derived from model representations. The experimental results demonstrate that Mix-MoE excels in multilingual MT, significantly outperforming existing baselines and showing notable progress in mitigatingparameter interference.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2605\.24681

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.24681 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2605.24681 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.24681 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Mixture of Experts (MoEs) in Transformers

Hugging Face Blog

Hugging Face blog post explaining Mixture of Experts (MoEs) architecture in Transformers, covering the shift from dense to sparse models, weight loading optimizations, expert parallelism, and training techniques for MoE-based language models.

MobileMoE: Scaling On-Device Mixture of Experts

Hugging Face Daily Papers

MobileMoE introduces efficient on-device mixture-of-experts language models with sub-billion parameters, achieving better performance and efficiency than dense baselines and existing MoE models. The models are trained on open-source datasets and demonstrate significant speedups on commodity smartphones.