Improving Multilingual Language Models by Aligning Representations through Steering

• URL: http://arxiv.org/abs/2505.12584v2
• Date: Tue, 26 Aug 2025 02:13:16 GMT
• Title: Improving Multilingual Language Models by Aligning Representations through Steering
• Authors: Omar Mahmoud, Buddhika Laknath Semage, Thommen George Karimpanal, Santu Rana,
• Abstract summary: This paper investigates how Large Language Models (LLMs) represent non-English tokens.<n>We propose a lightweight intervention method using representation steering, where a learned vector is added to the residual stream at a single model layer to enhance multilingual performance.
• Score: 10.159957091670883
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: This paper investigates how Large Language Models (LLMs) represent non-English tokens - a question that remains underexplored despite recent progress. We propose a lightweight intervention method using representation steering, where a learned vector is added to the residual stream at a single model layer to enhance multilingual performance. Through extensive experiments across seven competitive baselines -including prompt optimization, supervised fine-tuning (SFT), in-context learning, cross-lingual transfer, and translation-based methods-we show that our approach consistently outperforms most alternatives. In particular, it achieves performance on par with production-grade translation systems while requiring far fewer resources. We further explore the complementarity between our method and SFT, demonstrating that steering offers a direct, efficient way to realign internal representations. These findings underscore the potential of activation-level interventions as a powerful tool for improving the multilingual capabilities of LLMs.

Related papers

• CC-Tuning: A Cross-Lingual Connection Mechanism for Improving Joint Multilingual Supervised Fine-Tuning [48.69343479132896]
  CC-Tuning is a novel multilingual fine-tuning paradigm that explicitly establishes a cross-lingual connection mechanism at the latent level.<n>During training, CC-Tuning fuses the feed forward activations from both English and non-English inputs, enabling the model to benefit from both linguistic resources.<n>Experiments on six benchmarks covering 22 languages show that CC-Tuning outperforms vanilla SFT and offers a strong latent-level alternative to data-level augmentation methods.
  arXiv  Detail & Related papers  (2025-06-01T07:20:55Z)
• Semantic Aware Linear Transfer by Recycling Pre-trained Language Models for Cross-lingual Transfer [5.990773821761297]
  SALT is a novel cross-lingual transfer technique that recycles embeddings from target language Pre-trained Language Models (PLMs)<n>Our experiments show that SALT significantly outperforms other transfer methods and achieves lower loss with accelerating faster convergence during language adaptation.
  arXiv  Detail & Related papers  (2025-05-16T07:30:22Z)
• LayAlign: Enhancing Multilingual Reasoning in Large Language Models via Layer-Wise Adaptive Fusion and Alignment Strategy [33.85811169010525]
  Large language models (LLMs) exhibit suboptimal performance on low-resource languages.<n>Recent approaches have leveraged multilingual encoders alongside LLMs by introducing trainable parameters connecting the two models.<n>We propose aname, a framework that integrates representations from all encoder layers.
  arXiv  Detail & Related papers  (2025-02-17T03:45:03Z)
• Improving In-Context Learning with Small Language Model Ensembles [2.3499129784547654]
  In-context learning (ICL) is a cheap and efficient alternative but cannot match the accuracies of advanced methods.<n>We present Ensemble SuperICL, a novel approach that enhances ICL by leveraging the expertise of multiple fine-tuned small language models (SLMs)
  arXiv  Detail & Related papers  (2024-10-29T09:02:37Z)
• ShifCon: Enhancing Non-Dominant Language Capabilities with a Shift-based Multilingual Contrastive Framework [78.07201802874529]
  ShifCon is a Shift-based multilingual Contrastive framework that aligns the internal forward process of other languages toward that of the dominant one.<n>Experiments demonstrate that our ShifCon framework significantly enhances the performance of non-dominant languages.
  arXiv  Detail & Related papers  (2024-10-25T10:28:59Z)
• Bridging the Language Gaps in Large Language Models with Inference-Time Cross-Lingual Intervention [71.12193680015622]
  Large Language Models (LLMs) have shown remarkable capabilities in natural language processing. LLMs exhibit significant performance gaps among different languages. We propose Inference-Time Cross-Lingual Intervention (INCLINE) to overcome these limitations without incurring significant costs.
  arXiv  Detail & Related papers  (2024-10-16T11:23:03Z)
• EAGLE: Towards Efficient Arbitrary Referring Visual Prompts Comprehension for Multimodal Large Language Models [80.00303150568696]
  We propose a novel Multimodal Large Language Models (MLLM) that empowers comprehension of arbitrary referring visual prompts with less training efforts than existing approaches. Our approach embeds referring visual prompts as spatial concepts conveying specific spatial areas comprehensible to the MLLM. We also propose a Geometry-Agnostic Learning paradigm (GAL) to further disentangle the MLLM's region-level comprehension with the specific formats of referring visual prompts.
  arXiv  Detail & Related papers  (2024-09-25T08:22:00Z)
• FUSE-ing Language Models: Zero-Shot Adapter Discovery for Prompt Optimization Across Tokenizers [55.2480439325792]
  We propose FUSE, an approach to approximating an adapter layer that maps from one model's textual embedding space to another, even across different tokenizers. We show the efficacy of our approach via multi-objective optimization over vision-language and causal language models for image captioning and sentiment-based image captioning.
  arXiv  Detail & Related papers  (2024-08-09T02:16:37Z)
• ControlMLLM: Training-Free Visual Prompt Learning for Multimodal Large Language Models [73.34709921061928]
  We propose a training-free method to inject visual prompts into Multimodal Large Language Models (MLLMs)<n>We optimize a learnable latent variable based on an energy function, enhancing the strength of referring regions in the attention map.<n>Our method offers a promising direction for integrating referring abilities into MLLMs, and supports referring with box, mask, scribble and point.
  arXiv  Detail & Related papers  (2024-07-31T11:40:29Z)
• 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)
• Understanding the role of FFNs in driving multilingual behaviour in LLMs [0.0]
  In this paper, we conduct an in-depth analysis of the multilingual capabilities of a family of Large Language Models. We introduce novel metrics to probe the model's multilingual behaviour at different layers and shed light on the impact of architectural choices on multilingual processing.
  arXiv  Detail & Related papers  (2024-04-22T03:47:00Z)
• A Preference-driven Paradigm for Enhanced Translation with Large Language Models [33.51585908894444]
  Large language models (LLMs) can achieve remarkable translation performance using only a small amount of parallel data. SFT simply instructs the model to imitate the reference translations at the token level, making it vulnerable to the noise present in the references. We propose a preference-based approach built upon the Plackett-Luce model to overcome this plateau.
  arXiv  Detail & Related papers  (2024-04-17T11:52:47Z)
• Multi-modal Instruction Tuned LLMs with Fine-grained Visual Perception [63.03288425612792]
  We propose bfAnyRef, a general MLLM model that can generate pixel-wise object perceptions and natural language descriptions from multi-modality references. Our model achieves state-of-the-art results across multiple benchmarks, including diverse modality referring segmentation and region-level referring expression generation.
  arXiv  Detail & Related papers  (2024-03-05T13:45:46Z)
• Analyzing and Adapting Large Language Models for Few-Shot Multilingual NLU: Are We There Yet? [82.02076369811402]
  Supervised fine-tuning (SFT), supervised instruction tuning (SIT) and in-context learning (ICL) are three alternative, de facto standard approaches to few-shot learning. We present an extensive and systematic comparison of the three approaches, testing them on 6 high- and low-resource languages, three different NLU tasks, and a myriad of language and domain setups. Our observations show that supervised instruction tuning has the best trade-off between performance and resource requirements.
  arXiv  Detail & Related papers  (2024-03-04T10:48:13Z)
• Position-Enhanced Visual Instruction Tuning for Multimodal Large Language Models [50.07056960586183]
  We propose Position-enhanced Visual Instruction Tuning (PVIT) to extend the functionality of Multimodal Large Language Models (MLLMs) This integration promotes a more detailed comprehension of images for the MLLM. We present both quantitative experiments and qualitative analysis that demonstrate the superiority of the proposed model.
  arXiv  Detail & Related papers  (2023-08-25T15:33:47Z)
• DPL: Decoupled Prompt Learning for Vision-Language Models [41.90997623029582]
  We propose a new method, Decoupled Prompt Learning, which reformulates the attention in prompt learning to alleviate this problem. Our approach is flexible for both visual and textual modalities, making it easily extendable to multi-modal prompt learning.
  arXiv  Detail & Related papers  (2023-08-19T15:48:38Z)
• Optimal Transport Posterior Alignment for Cross-lingual Semantic Parsing [68.47787275021567]
  Cross-lingual semantic parsing transfers parsing capability from a high-resource language (e.g., English) to low-resource languages with scarce training data. We propose a new approach to cross-lingual semantic parsing by explicitly minimizing cross-lingual divergence between latent variables using Optimal Transport.
  arXiv  Detail & Related papers  (2023-07-09T04:52:31Z)
• On the cross-lingual transferability of multilingual prototypical models across NLU tasks [2.44288434255221]
  Supervised deep learning-based approaches have been applied to task-oriented dialog and have proven to be effective for limited domain and language applications. In practice, these approaches suffer from the drawbacks of domain-driven design and under-resourced languages. This article proposes to investigate the cross-lingual transferability of using synergistically few-shot learning with prototypical neural networks and multilingual Transformers-based models.
  arXiv  Detail & Related papers  (2022-07-19T09:55:04Z)
• Incorporating Linguistic Knowledge for Abstractive Multi-document Summarization [20.572283625521784]
  We develop a neural network based abstractive multi-document summarization (MDS) model. We process the dependency information into the linguistic-guided attention mechanism. With the help of linguistic signals, sentence-level relations can be correctly captured.
  arXiv  Detail & Related papers  (2021-09-23T08:13:35Z)
• SLM: Learning a Discourse Language Representation with Sentence Unshuffling [53.42814722621715]
  We introduce Sentence-level Language Modeling, a new pre-training objective for learning a discourse language representation. We show that this feature of our model improves the performance of the original BERT by large margins.
  arXiv  Detail & Related papers  (2020-10-30T13:33:41Z)

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.