On the Limits of Model Merging for Multilinguality in Pre-Training

• URL: http://arxiv.org/abs/2605.25846v1
• Date: Mon, 25 May 2026 13:38:33 GMT
• Title: On the Limits of Model Merging for Multilinguality in Pre-Training
• Abstract summary: We conduct a controlled study on the efficacy of mixed, merged, and monolingual pre-training setups.<n>We find that while monolingual pre-training results in strong in-language performance, merging any combination of monolingual models leads to performance collapse due to interference.
• Score: 49.9160455376296
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Endowing models with consistent multilingual performance can be achieved by mixing pre-training data, or post-training approaches such as language-specific model merging. In this work, we test whether merging can be applied to monolingually pre-trained models. We conduct a controlled study on the efficacy of mixed, merged, and monolingual pre-training setups. We find that while monolingual pre-training results in strong in-language performance, merging any combination of monolingual models leads to performance collapse due to interference. Our analysis suggests representational similarity is a prerequisite for model merging. We therefore conclude that the flexibility of merging in fine-tuning does not extend trivially to language-specific pre-training.

Related papers

• Distilling Monolingual and Crosslingual Word-in-Context Representations [18.87665111304974]
  We propose a method that distils representations of word meaning in context from a pre-trained language model in both monolingual and crosslingual settings. Our method does not require human-annotated corpora nor updates of the parameters of the pre-trained model. Our method learns to combine the outputs of different hidden layers of the pre-trained model using self-attention.
  arXiv  Detail & Related papers  (2024-09-13T11:10:16Z)
• PreAlign: Boosting Cross-Lingual Transfer by Early Establishment of Multilingual Alignment [68.20851615263953]
  Large language models demonstrate reasonable multilingual abilities, despite predominantly English-centric pretraining. The spontaneous multilingual alignment in these models is shown to be weak, leading to unsatisfactory cross-lingual transfer and knowledge sharing. We propose PreAlign, a framework that establishes multilingual alignment prior to language model pretraining.
  arXiv  Detail & Related papers  (2024-07-23T06:59:53Z)
• Improving In-context Learning of Multilingual Generative Language Models with Cross-lingual Alignment [42.624862172666624]
  We propose a simple yet effective cross-lingual alignment framework exploiting pairs of translation sentences. It aligns the internal sentence representations across different languages via multilingual contrastive learning. Experimental results show that even with less than 0.1 textperthousand of pre-training tokens, our alignment framework significantly boosts the cross-lingual abilities of generative language models.
  arXiv  Detail & Related papers  (2023-11-14T11:24:08Z)
• Analyzing the Mono- and Cross-Lingual Pretraining Dynamics of Multilingual Language Models [73.11488464916668]
  This study investigates the dynamics of the multilingual pretraining process. We probe checkpoints taken from throughout XLM-R pretraining, using a suite of linguistic tasks. Our analysis shows that the model achieves high in-language performance early on, with lower-level linguistic skills acquired before more complex ones.
  arXiv  Detail & Related papers  (2022-05-24T03:35:00Z)
• Lifting the Curse of Multilinguality by Pre-training Modular Transformers [72.46919537293068]
  multilingual pre-trained models suffer from the curse of multilinguality, which causes per-language performance to drop as they cover more languages. We introduce language-specific modules, which allows us to grow the total capacity of the model, while keeping the total number of trainable parameters per language constant. Our approach enables adding languages post-hoc with no measurable drop in performance, no longer limiting the model usage to the set of pre-trained languages.
  arXiv  Detail & Related papers  (2022-05-12T17:59:56Z)
• On the ability of monolingual models to learn language-agnostic representations [2.604227467422371]
  We show that monolingual models pretrained and finetuned on different languages achieve competitive performance. For example, models pretrained on distant languages such as German and Portuguese perform similarly on English tasks.
  arXiv  Detail & Related papers  (2021-09-04T22:09:44Z)
• ERNIE-M: Enhanced Multilingual Representation by Aligning Cross-lingual Semantics with Monolingual Corpora [21.78571365050787]
  ERNIE-M is a new training method that encourages the model to align the representation of multiple languages with monolingual corpora. We generate pseudo-parallel sentences pairs on a monolingual corpus to enable the learning of semantic alignment between different languages. Experimental results show that ERNIE-M outperforms existing cross-lingual models and delivers new state-of-the-art results on various cross-lingual downstream tasks.
  arXiv  Detail & Related papers  (2020-12-31T15:52:27Z)
• How Good is Your Tokenizer? On the Monolingual Performance of Multilingual Language Models [96.32118305166412]
  We study a set of nine typologically diverse languages with readily available pretrained monolingual models on a set of five diverse monolingual downstream tasks. We find that languages which are adequately represented in the multilingual model's vocabulary exhibit negligible performance decreases over their monolingual counterparts.
  arXiv  Detail & Related papers  (2020-12-31T14:11:00Z)
• Mixed-Lingual Pre-training for Cross-lingual Summarization [54.4823498438831]
  Cross-lingual Summarization aims at producing a summary in the target language for an article in the source language. We propose a solution based on mixed-lingual pre-training that leverages both cross-lingual tasks like translation and monolingual tasks like masked language models. Our model achieves an improvement of 2.82 (English to Chinese) and 1.15 (Chinese to English) ROUGE-1 scores over state-of-the-art results.
  arXiv  Detail & Related papers  (2020-10-18T00:21:53Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.