Examining Modularity in Multilingual LMs via Language-Specialized Subnetworks

• URL: http://arxiv.org/abs/2311.08273v1
• Date: Tue, 14 Nov 2023 16:11:23 GMT
• Title: Examining Modularity in Multilingual LMs via Language-Specialized Subnetworks
• Abstract summary: We investigate the degree to which language-wise modularity naturally arises within models with no special modularity interventions. We show that language-specializedworks do naturally arise, and that SFT, rather than always increasing modularity, can decrease language specialization of Attributionworks in favor of more cross-lingual sharing.
• Score: 15.758213562260913
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recent work has proposed explicitly inducing language-wise modularity in multilingual LMs via sparse fine-tuning (SFT) on per-language subnetworks as a means of better guiding cross-lingual sharing. In this work, we investigate (1) the degree to which language-wise modularity naturally arises within models with no special modularity interventions, and (2) how cross-lingual sharing and interference differ between such models and those with explicit SFT-guided subnetwork modularity. To quantify language specialization and cross-lingual interaction, we use a Training Data Attribution method that estimates the degree to which a model's predictions are influenced by in-language or cross-language training examples. Our results show that language-specialized subnetworks do naturally arise, and that SFT, rather than always increasing modularity, can decrease language specialization of subnetworks in favor of more cross-lingual sharing.

Related papers

• MoLGE: Mixture of Language Group Experts for Efficient Scaling of Massively Multilingual Speech Recognition [23.519328539637794]
  MoLGE assigns dedicated expert modules to clusters of similar languages.<n>It integrates a hierarchical Low-Rank Adaptation (LoRA) strategy into the disentangled acoustic and linguistic components of the S3M architecture.<n>Results demonstrate that MoLGE consistently outperforms dense multilingual baselines with a minimal increase in trainable parameters.
  arXiv  Detail & Related papers  (2026-07-27T05:55:50Z)
• Leveraging Routing Dynamics in Mixture-of-Experts Models for Efficient Language Adaptation [50.731838216268706]
  We study multilingual routing dynamics during continual pre-training of an English-centric MoE model on a multilingual corpus.<n>We find that continual multilingual pre-training leads to diffused, language-agnostic routing in early and middle layers.<n>We propose a parameter-efficient adaptation strategy that updates language-specific and shared experts in the final MoE layers.
  arXiv  Detail & Related papers  (2026-05-28T10:12:41Z)
• Adapting Language Models to Indonesian Local Languages: An Empirical Study of Language Transferability on Zero-Shot Settings [1.1556013985948772]
  We evaluate transferability of pre-trained language models to low-resource Indonesian local languages.<n>We group the target languages into three categories: seen, partially seen, and unseen.<n> Multilingual models perform best on seen languages, moderately on partially seen ones, and poorly on unseen languages.<n>We find that MAD-X significantly improves performance, especially for seen and partially seen languages, without requiring labeled data in the target language.
  arXiv  Detail & Related papers  (2025-07-02T12:17:55Z)
• The Unreasonable Effectiveness of Model Merging for Cross-Lingual Transfer in LLMs [54.59207567677249]
  Large language models (LLMs) still struggle across tasks outside of high-resource languages.<n>In this work, we investigate cross-lingual transfer to lower-resource languages where task-specific post-training data is scarce.
  arXiv  Detail & Related papers  (2025-05-23T20:28:31Z)
• High-Dimensional Interlingual Representations of Large Language Models [65.77317753001954]
  Large language models (LLMs) trained on massive multilingual datasets hint at the formation of interlingual constructs. We explore 31 diverse languages varying on their resource-levels, typologies, and geographical regions. We find that multilingual LLMs exhibit inconsistent cross-lingual alignments.
  arXiv  Detail & Related papers  (2025-03-14T10:39:27Z)
• MoE-CT: A Novel Approach For Large Language Models Training With Resistance To Catastrophic Forgetting [53.77590764277568]
  We introduce a novel MoE-CT architecture that separates the base model's learning from the multilingual expansion process. Our design freezes the original LLM parameters, thus safeguarding its performance in high-resource languages, while an appended MoE module, trained on diverse language datasets, augments low-resource language proficiency.
  arXiv  Detail & Related papers  (2024-06-25T11:03:45Z)
• Soft Language Clustering for Multilingual Model Pre-training [57.18058739931463]
  We propose XLM-P, which contextually retrieves prompts as flexible guidance for encoding instances conditionally. Our XLM-P enables (1) lightweight modeling of language-invariant and language-specific knowledge across languages, and (2) easy integration with other multilingual pre-training methods.
  arXiv  Detail & Related papers  (2023-06-13T08:08:08Z)
• High-resource Language-specific Training for Multilingual Neural Machine Translation [109.31892935605192]
  We propose the multilingual translation model with the high-resource language-specific training (HLT-MT) to alleviate the negative interference. Specifically, we first train the multilingual model only with the high-resource pairs and select the language-specific modules at the top of the decoder. HLT-MT is further trained on all available corpora to transfer knowledge from high-resource languages to low-resource languages.
  arXiv  Detail & Related papers  (2022-07-11T14:33:13Z)
• Discovering Language-neutral Sub-networks in Multilingual Language Models [15.94622051535847]
  Language neutrality of multilingual models is a function of the overlap between language-encoding sub-networks of these models. Using mBERT as a foundation, we employ the lottery ticket hypothesis to discover sub-networks that are individually optimized for various languages and tasks. We conclude that mBERT is comprised of a language-neutral sub-network shared among many languages, along with multiple ancillary language-specific sub-networks.
  arXiv  Detail & Related papers  (2022-05-25T11:35:41Z)
• 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)
• Multilingual Neural Machine Translation:Can Linguistic Hierarchies Help? [29.01386302441015]
  Multilingual Neural Machine Translation (MNMT) trains a single NMT model that supports translation between multiple languages. The performance of an MNMT model is highly dependent on the type of languages used in training, as transferring knowledge from a diverse set of languages degrades the translation performance due to negative transfer. We propose a Hierarchical Knowledge Distillation (HKD) approach for MNMT which capitalises on language groups generated according to typological features and phylogeny of languages to overcome the issue of negative transfer.
  arXiv  Detail & Related papers  (2021-10-15T02:31:48Z)
• Learning Multilingual Representation for Natural Language Understanding with Enhanced Cross-Lingual Supervision [42.724921817550516]
  We propose a network named decomposed attention (DA) as a replacement of MA. The DA consists of an intra-lingual attention (IA) and a cross-lingual attention (CA), which model intralingual and cross-lingual supervisions respectively. Experiments on various cross-lingual natural language understanding tasks show that the proposed architecture and learning strategy significantly improve the model's cross-lingual transferability.
  arXiv  Detail & Related papers  (2021-06-09T16:12:13Z)
• Adaptive Sparse Transformer for Multilingual Translation [18.017674093519332]
  A known challenge of multilingual models is the negative language interference. We propose an adaptive and sparse architecture for multilingual modeling. Our model outperforms strong baselines in terms of translation quality without increasing the inference cost.
  arXiv  Detail & Related papers  (2021-04-15T10:31:07Z)
• Improving the Lexical Ability of Pretrained Language Models for Unsupervised Neural Machine Translation [127.81351683335143]
  Cross-lingual pretraining requires models to align the lexical- and high-level representations of the two languages. Previous research has shown that this is because the representations are not sufficiently aligned. In this paper, we enhance the bilingual masked language model pretraining with lexical-level information by using type-level cross-lingual subword embeddings.
  arXiv  Detail & Related papers  (2021-03-18T21:17:58Z)
• VECO: Variable and Flexible Cross-lingual Pre-training for Language Understanding and Generation [77.82373082024934]
  We plug a cross-attention module into the Transformer encoder to explicitly build the interdependence between languages. It can effectively avoid the degeneration of predicting masked words only conditioned on the context in its own language. The proposed cross-lingual model delivers new state-of-the-art results on various cross-lingual understanding tasks of the XTREME benchmark.
  arXiv  Detail & Related papers  (2020-10-30T03:41:38Z)

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.